Architecture is becoming more reflective. Not because architects are theorizing more about what they do – they have always done this –, but because the conditions of the possibility of architecture are themselves becoming an object of architectural design. Stan Allen’s post-semiotic architecture of infrastructures is an example. According to Allen (Infrastructural Ubanism), “Under the dominance of the representational model, architecture has surrendered its capacity to imagine, to propose, or to construct alternative realities.” (50)  This implies a new program for architecture in which design does not exhaust itself in autonomous, representative buildings, but is concerned primarily with “the production of directed fields in which program, event, and activity can play themselves out.” (52) Concretely, Allen is talking about infrastructures, that is, not specific buildings, but “the site itself…the conditions for future events…the construction of surfaces, the provision of services to support future programs, …the establishment of networks for movement, communication, and  exchange.” (54) All these things such as transportation systems, energy grids, communication networks, and so on can be seen as the conditions of the possibility for architecture in the sense of objects and buildings. For readers of Castells’ Information Society this sounds a lot like what Castells calls the “space of flows,” that is, the hardware and software that creates a global network of simultaneous action beyond any local places, traditions, cultures, and identities. Although Castells has not entirely given up the conflict and even contradiction between the global and somewhat virtual space of flows on the one side and the physical and fragmented space of places, he has come to recognize the role of architecture in “making places in the space of flows” (Space of Flows, Space of Places: Materials for a Theory of Urbanism in the Information Age, in The Cybercities Reader ed. S. Graham, 2004). Castells cites such projects as Gehry’s Guggenheim Museum in Bilbao, Calatrava’s bridges, telecommunication towers, airports, Moneo’s AVE station in Madrid, and Koolhaas’s Lille Grand Palais. Many other multipurpose megastructures in the form of transportation or communication hubs could be added to the list. In these structures what is designed is infrastructure, the network of the network society. It is not place but flow that is the “object” of architectural design. It is the flow that becomes the place we live, work, play, shop, etc. Contrary to Castells opposition between flow and place, what these examples show is that it is the space of flows that is becoming the place in which we live, work, and construct our identities.

This is the thesis of architect and philosopher Gilles Delalex’s interesting book Go with the Flow (U. of Art & Design Helsinki, 2006). Delelax (founder of MUOTO Architects) is well known for his infrastructural projects, above all, in association with motorways. With regard to the problem of bridging the gap between the space of flows and the space of places, we think Delelax should be given an award for asking all the right questions. Here are some them:

“How do global flows solidify into urban topographies?”

“What is the role of architecture in the space of flows?”

“Can flows and buildings be conceived of as a mutual construction of each other?”

Delelax claims that “in the fields of architecture and urban design, the issue is not therefore to design places or objects, but the flows that lead from one to another.” (15) This means that the space of flows “provides a new paradigm for architecture and urban design” (59). And since it is infrastructures that enable, condition, steer, channel, and modulate flows of all kinds, whether it be people, automobiles, money, goods, airplanes, information, or anything else, Delelax follows Allen in seeing the task of architecture in the design of infrastructure. But what is infrastructure?

Following Castells’ definition of flow, Delelax defines infrastructure in terms of processes of repetition, programming, exchange, interaction, and simultaneity.  This includes both hardware (physical objects such as roads, buildings, pipelines, etc.) and software (programs for use, meanings, opportunities for action, etc.).  Places are not historically given, unchangeable, and therefore marginalized fragments existing somehow outside the global connectivity and flow of the network society. Against Castells, Delelax understands places as what emerges from flows that are conditioned by infrastructures. The role of the infrastructure and thus the task of architecture is to act as a kind of filter for the emergence of places. What kind of place we are at any time in and the kinds of information and activities that are in any particular place possible, are a result of the filtering function of the infrastructure. Mulitpurpose megastructures are complex filters that let certain kinds of activities, such as work, shopping, entertainment, travel, living, health care, education, cultural performances, etc. happen or not. They let certain places emerge from the programs they support and hinder other places from appearing by hindering the flows of information and opportunities required by these places. Even in the space of flows not everything is possible all the time at the same place. As Delelax puts it “The question is no longer what objects and buildings are made out of, but what they interface with.” (15) And this in turn means that “The space of lows entails today a conception of space that marks a shift from the modern Cartesian space, where objects were geometrically organized on a neutral and horizontal background, to a more virtual space defined by intricate relations between cities, companies or individuals. It evokes a space which is no longer understood as a neutral ground (of representation or intervention), but as an active subject, agent or process.” (15)

One way of understanding infrastructure and thus architecture as an agent is to focus on its function as filter, as gateway, as hub in a network that channels, steers, includes and excludes information and action. This perspective brings architecture near to what interface is, namely, a filter for knowledge and action.  Vice versa, architecture as infrastructure permits a new understanding of interface. Interfaces are no longer screens or input devices for access to computers. Instead, the interface becomes the place of flow, the form of space in the network society.