Still from You, the Living by Roy Andersson (2007)
Text originally published as “Shelter’s Political Violence,” in Arjen Oosterman and Nick Axel (eds), Shelter, Volume Magazine 46 (2015)
Shelter is often interpreted as the original typology of architecture, both for its primitiveness and for the simplicity of its function: a shelter is an architecture that protects the bodies it hosts from external conditions, such as the rain, the wind, the snow, etc. However, as often when it comes to the original figure of something, there is a risk of transforming this figure into a ‘pre-political’ myth, which recalls the illusory opposition between Jean-Jacques Rousseau and Thomas Hobbes seeking to define the natural condition of the object we consider. The question is not whether the pre-political subject is intrinsically good (Rousseau) or bad (Hobbes) but, rather, if this very notion of pre-politics exists and helps us to interpret the world. Our imaginary tends to confirm this pre-political vision: after all, what is more innocent than a simple shelter whose only function is to protect our bodies from the elements? By extension, if we were to prove such innocence, we could draw the conclusion that architecture does not always bear political consequences. To the contrary, this text will attempt to demonstrate that the primitive shelter contains the totality of architectural characteristics that make the organization of space violent, regardless of the intentions that motivated its construction.
Let’s try to consider a shelter, the bodies it hosts and its direct environment, in utter simplicity: whether it consists of simply a roof or a walled enclosure of some sort, there is a separation – either an abrupt or a gradual one – between the sheltered space and the space outside of it. The sheltered space is, by definition, limited, which means that it can only contain a finite amount of matter. Bodies are material assemblages, which means that a shelter can only contain a limited amount of bodies and, by extension, that only a limited amount of bodies will be sheltered.
The protocol that defines which bodies get to be sheltered and which do not varies depending on a number of factors – architecture itself being a significant one. For the sake of this demonstration, let’s not yet assume that any specific politics would have driven its design. This does not necessarily prevent a protocol of selection to be applied, even if it is based on such a trivial rule as ‘first come first served’. We can easily picture a roofed space during heavy rain filling up with bodies seeking refuge. No other protocol applies here than the simple progressive accumulation of bodies under a limited roof (assuming that a physical antagonism will not develop between bodies). Eventually, the moment comes when the critical threshold is reached and any additional body seeking shelter simply cannot access the protection offered under the roof due to the dense population of bodies already there present. It is too early to apply a judgment on this de facto exclusion of a body – is it fair? is it ethical? is it legal? – so let us simply observe that this exclusion occurs, and that it occurs because of an engineered separation of a sheltered space from a non-sheltered space. This is the crucial point of the argument I am attempting to make: the status of the non-sheltered body is defined by the negativity of its exclusion as a direct opposite to the sheltered body. As we can see through the very name of their status, what socially separates these bodies is architecture itself: shelter.
“It’s just rain!” one might say, and the politics of who gets wet and who doesn’t still appears relatively inconsequential. What is true for rain is however also true for all other external conditions applied to a given body. Shelter provides protection against the violence of these conditions, and it is rare that its protocol of differentiating bodies that can access such protection and those that cannot is simply based on the ‘first come, first served’ rule. More often than not, the protocol of shelter refers to the rules defined by the regime of private property that no longer grants bodies access to protection based on how many can be accommodated, but rather on a more drastic political selection through acceptability. The criteria of such acceptability are multiple and various, but they rarely escape from the normative categorization of bodies into race, gender, sexual preference, ability, etc. and the violence it implies. For example, we can think of the selective criteria for accepting refugees currently coming from Afghanistan, Syria or Eritrea who seek refuge in Western, European countries. Many political representatives have expressed their will to select refugees to accommodate based on religion (Christians, rather than Muslims) and gender (women, rather than men); thus a protocol of access to a particular form of shelter is formed. It ought to be noted that shelter operates at several scales and being granted access to ‘national shelter’ does not necessarily prevent one from being excluded by other local forms of shelter and their respective protocols of access.
To simplify this demonstration and to show that architecture does not need to be complex to become both applicable and operative, let’s return to where we started, with the example of a simple roof. Such an architectural organization of space is particularly effective to accommodate a ‘first come, first serve’ protocol, however architecture can become more complex to accommodate more drastic protocols of selection, like the ones mentioned above. As material assemblages, walls are largely constructed in such a way that their structural integrity cannot be affected by bodies without proper tools that are themselves subject to the protocol of private property, acceptability and access. Hence why an imprisoned body will not succeed in evading the walls that surround it, yet the buildings experienced by our bodies on a daily basis are not essentially different; one cannot simply cross walls to transgress the way space has been designed and bodies separated. The difference between our homes and prison cells can be found in the ability for a body to directly intervene into the walls’ porosity through architectural devices, such as doors and windows, to which the protocol of admission has been materialized in a small object called the key. Who owns the key is a body included as acceptable in the protocol of admission defined by both a legal regime (private property for instance), and a material one (the wall and its porosity regulator, the door).
In French, homeless bodies are called shelter-less (sans-abri). They are the bodies not included in any selective protocol that grants access to a shelter; they are key-less, and all walls are reminders of the outdoor prison from which they cannot escape. This notion of the ‘outdoor prison’ may sound exaggerated; after all, the exterior milieu of the shelter-less is what remains in the world after we have withdrawn the totality of the sheltered spaces from it. However, I would like to insist on the fact that this ‘remaining world’ does not constitute an untouched condition in which bodies would have the same status as one in which shelters had never been erected. In other words, shelter (and, by extension, architecture) does not only create the status of the selected and protected body; it also actively constructs the excluded and shelter-less one. There are two sides to each wall, regardless of which one may be covered by a roof and which one may not.
Shelter, whatever its form and intention, might indeed incarnate the archetypical architectural typology, but it does so through the inherent political violence of segregation and exclusion it deploys. Architects and designers should not shy away from this violence, but rather understand its inexorability as well as its modus operandi in order to use it to inform their own politics, whatever they might be. Only a political manifesto that conforms with all points with the normative set of power relations at a given moment can enjoy the luxury of not voluntarily adopting a position, since a ‘non-political’ architecture will necessarily reinforce the current dominant politics. Designing architecture means also designing the protocols of its access. If we want these protocols to conform to the ubiquitous relationships of power proposed by the norm, we can surely pretend to ignore this violence. If, on the other hand, we want to challenge these relationships, we need to appropriate architecture’s inherent violence to reconfigure the way bodies are organized in space; to condition – architecture cannot pretend to do more than this – other forms of social relationships undermining the normative configuration of society that determine the various protocols of acceptability.