WHY: Theoretical Framework

Version 1.4

If you have not read the "Production Concept" and the "Project Background" pages, they will be useful before reading further.  The production concept establishes the method of intervention, the how of the material premise that grounds the rest of the project -- much as an artistic intention must be tethered to a specific material like oil paint, colored light, or stone.  The specifics of the form of the intervention will be addressed over the next several months as the work progresses, as recorded in the "WHAT" blog feed.  Lastly, there is the more fundamental question of the nature of this intervention – the why of it.  This will be addressed below, to set the framework for the nature and the necessity of the intervention being proposed.   

Entry Point 

As described by the production page, the architectural concept being presented involves the resurrection of a long dormant form of architecture:  rock-cut architecture.  And, as explained in the project background, this is specifically as a response to an open-ended question of possibilities for how architecture can address sustainability.  The forms and patterns of our built environment are critical focal points for questions of sustainability. However, hopefully obviously, this proposal is not a general concept to revive rock-cut architecture as a material means of producing our cities.  Rather it is for the production of one very specific piece of architecture that is currently absent.  The framework that follows will, if successful, clarify this absence and provide a case for rock-cut architecture as the necessary response.

The typical formulations of what architecture can offer our efforts to address sustainability involve energy efficiency, recycled building materials, renewable energy sources, life cycle costs, material embodied energy, urban density and economies of scale, revised patterns of habitation and transportation, spaces for localized production of necessities like potable water, gardening, etc.  All these issues are and will be important to address, however these bracketed lines of problem-solving respond to questions that are structured by the conceptual foundations of modern architecture, where possibility is defined by a functional positivism that aspires only toward system optimization.  There are however, other forms of architectural thought which start from different kinds of questions -- the history of architecture demonstrates many -- and that lead to architectural possibilities that are difficult destinations to access through the logic of modern architecture.  The architecture of this proposal is one such possibility that falls outside of modern architectural thought. 

An Exit Through Actor-Network Theory

We can't solve problems by using the same kind of thinking we used when we created them.   


To open this question that leads one to an answer of “rock-cut architecture,” we have to place some speed bumps along the path of our head-long dash toward clarity of understanding and subsequent solution-seeking machinations.  We will do this using two almost completely unrelated sources -- Actor-Network Theory (ANT), and Heidegger -- chosen for their ability to target foundational features of our everyday thinking.  The goal here is not strict scholarly fidelity or academic compliance, but to play on their ability to momentarily suspend our most fundamental ready-to-hand habits of (architectural) thought – an essential analytical step within the larger synthetic goal of building the conceptual foundation of this proposal’s design intervention.

Actor-Network Theory offers a useful conceptual entry point for reflections on sustainability due to its open-ended inclusiveness, in that it offers a framework for mapping the networks of complex, generative relationships between humans and non-humans that compose our world.  Firstly, to the degree that ANT retains a foundational ontology at all, it is premised on a revision: essence is networked.  By all canonized accounts, modern ontology assumes the spatiality of Being – three (at least) dimensionally extended objects, distributed in space and ranging in scale, subjects in space perceiving the objects, etc.  The ontology proposed by ANT supplants this spatial foundation with the network, which is by nature a-spatial.  Spatiality gives way to relationality.  As a methodology, ANT is a hybrid material-semiotic mode of description.  It offers no explanation, prediction, or normative claims.  It offers only a toolkit for mapping its own redefined version of ontology.  It is to networks what cartography is to a planet – in the sense that the value offered by both is the same: enabling a mode of accurate description opens new possibilities of engagement and intervention.  By rendering networks of relationships explicit, ANT makes these relationships possible sites of manipulation.  ANT maps the networks of dependencies, influences, and exchanges that cohere humans and non-humans into temporarily stablized systems that produce specific outcomes; whether a fact, a technology, or a collectively arrived at decision – with the additional claim that it is the relationships shown in these network maps that anchor reality, not the spatial terrain we inhabit and the objects therein.  The real is an unstable constellation of temporarily stable production networks, and ANT is the mode of revelation.

ANT networks are composed of two basic parts: 1) actors are the equivalent of network nodes that perform some action in the network, and 2) a web of “unspecific relationships”, which in ANT terminology translates the actors into a coherent network. 

As mentioned, ANT is a mode of material-semiotic description, signalling the divide between semiotics/textuality/discourse on the one side and objective materiality on the other, a divide which according to ANT is erased through revealing hybrid relationships composed of features from both ends of this conflicted dichotomy, i.e. if properly conceived there is no actual division.  These heterogeneous networks cannot be understood solely through semiotics without the referent of science, and vice versa; just as the social cannot be isolated from the material practices that feed it (would the “semiotic turn” have been possible without the printing press?), and material practices cannot emerge without the social/semiotic processes that order them: they are different aspects of the same thing, revealed through a relational network.  In what amounts to a sort of concession to the human origin of an ANT analysis, the starting point of ANT is that all the actors in a network, human or non-human, are semiotically constructed.  ANT strategically minimizes initial assumptions to maximize its freedom of application.  Actor is effectively a placeholder concept, a container that can hold anything – it assumes no properties, distinctions, axioms, associations, scale, location, etc.  Everything attributable to an actor is assigned through its semiotic construction during the ANT analysis, meaning the placeholder is only filled with a definition to the degree required by its function within the network.  Yet, the natural/objective/material world is the category of actor that cannot not be constructed for a network to be coherent.

The second part of an ANT network, the connections, are simply defined as the path of what it calls an intermediary, which is anything that carries action between parts of a network: a text, an object, an idea, research results, data, information, force, speech, capacity, etc.  For example, if we wanted to describe changes to bacteria populations over the past 200 years, what spectrum of actors would require attention?  A scientific practice that semiotically constructed bacteria into a feature of our world, a technological practice that developed anti-bacterial drugs/poisons to target them for intervention and control, the medical/institutional material practices of anti-bacterial drug usage, the economics of consumer-generated market pressures to expand their scope of use, the bacterial response of accelerated mutation in the presence of our antibacterial interventions, a progressive re-shaping of the proportions and diversity of the bacterial biomass, scientific research studying the consequences of our material practices, the social engineering of counter-messaging about over-usage to reduce market pressures, and on and on.  This heterogeneous tangle of actors and intermediaries collectively constitute a web of interdependency spanning between the objective world, through technology and engineering, to the social production of knowledge and habits – each facet cannot be understood without the others, and they have collectively stabilized into a network exchange of data, policy, texts, technologies, funding, causality, materials, etc.  It is the breadth of necessary network inclusions, in term of actors and intermediaries, that make necessary the minimal initial architecture of ANT, only to be expanded, diversified, and articulated through a methodical but open-ended inquiry.  However, the extent of the network to be analyzed, or how far the network map is traced, is truncated at a certain point as a matter of either convenience or exhaustion for the ANT analyst, but never due to completion.  The tracing of a network extends until the actants that have yet to be mapped into the network as actors are assumed to be of negligible consequence on the output.  The threshold of “negligible” is at the discretion of the ANT researcher.  

All these disparate actors are mapped in a homogenous fashion, and most controversially, all assigned various degrees of the same kind of agency through its law of generalized symmetry.  ANT dismisses a deeply ingrained belief that humans are unique, making us a discreet category from the rest of what is in the world.  ANT is not resurrecting some form of animism, extending human agency to all the rest, but splitting the difference with the inverse.  It is not simply a promotion of non-humans, but also a demotion of humans, settling somewhere in between.  According to ANT, human agency is a decentralized network effect, not different in kind from the rest of the actors in a network, with each actor having agency to the degree that it impacts the network.  There is little room within ANT analysis for subjectivity, being one of the features in our auto-narration of the world that recedes when remade in network terms.

Most importantly, ANT reconfigures our concept of knowledge.  In modern epistemology knowledge is binary, true or false, as measured against the objective world.  In ANT, truth is measured against what amounts to “networks of alliance.”  In the playing field of knowledge competing against rival knowledge, the game is not won by “accuracy of correlation to reality,” but by rallying networks of allies.  ANT can be thought of as Google’s Pagerank (the search algorithm that ranks returns) elevated to the level of epistemological yardstick: truth is a matter of degree with the “most true” fact being the most connected, referenced, cited, and sought fact.  Like human agency, conventional concepts like “power” and “truth” are redefined as network effects that solely result from the expanse of a network and an actor’s position/function within it.   These networks span the full range of “actors” associated with producing knowledge, from researchers, institutions, and funding sources to technologies, devices, ideas, objects of study, and the ever-crucial: community of other researchers.  However, it should be emphasized for our purpose here, that this conceptual framework applies to collective processes that produce decisions as much as to those that produce knowledge, i.e. a treaty is as mappable as a research finding.  

Space and geography are demoted to just another possible inclusion in a network.  ANT is fond of claiming that it eliminates space and scale:  I am far closer to a colleague in another country than to my neighbor that I never speak to, which would be reflected in a network map that contained the three of us.  The relational network connecting me and the colleague is more fundamental than the spatial proximity between me and the neighbor.  And in network terms, scale is meaningless – the only measure of a virus versus a human is the force it can potentially exert within a network, relative physical size is irrelevant except for whatever potential effects this could have on the extent of the network, for example the necessity of developing an intermediate technology to extend human perception to “see” the virus.  By supplanting space, scale, and geography with relationality, ANT effectively positions itself as the antithesis of architecture – which is precisely what makes it an appropriate counter-balance to our ingrained paths of architectural thought.  What is the import of architecture in a world rebuilt in network terms?

Actor-Network Theory, Environmentalism, and the Architecture of this Proposal

A comprehensive ANT analysis of any given feature of “the environmental movement" is far beyond the scope of this project.  However, when our initial premise at the intersection of environmentalism and architecture is viewed through the ANT framework, some cursory results surface:  

1.     The opening question of this project was addressed to architecture, not building envelope technology, or mechanical engineering, or building material science, etc.  The question asks what a work of architecture can offer, as opposed to possible additions to the list of sustainable features to be implemented in the built fabric-at-large.  Architecture, as an artifact, is spatially localized.  It cannot be mediated or transferred as with images, music, text, sculpture, etc.  Its scale, site specificity, and its prerequisite of an experiencing body in space, bracket what kind of environmental leverage points can be hit using architecture as the medium.  In the terms of an ANT map, the “environmentalism network" would reveal two kinds of leverage points:  1) diffuse, spread across the patterns of the general populace, and 2) concentrated, which correlate with centers of power, knowledge production, and decision-making.  As follows from architecture's spatial fixity, the nature of a single work of architecture lends itself to the latter set of leverage points.  The ANT/environmentalism network map would increase in density of nodes and connections toward the highest leverage point(s), inevitably arriving at international agreements on climate targets and the processes that produce them.  The environmental science research institutes producing the “hard data” of the environmentalism debate are each a network (also themselves composed of “sub”-networks of programs, labs, etc), which collectively form a “scientific community” network, that is again part of a larger network that includes governments and all the other players involved in international climate summits.  Each of these levels results in an emergent form of knowledge and/or decision production that contribute toward the output at the highest leverage point, the organizational “top”, the international forums that produce global accords, treaties, etc.-- the network concentration points which take place within space provided by architecture.  The time and space in which these highest leverage points occur are highly specific, meaning a single work of architecture can contain them.  The highest environmental leverage point that architecture can address is the space in which global environmental summits occur.

2.     In the process of analyzing the production of a global climate treaty, ANT methodology forces the inquiry to open far beyond the people involved and the interests they represent – not only in terms of the chain of sub-networks that contribute to the process, but more importantly it clears a conceptual space in which the as of yet latent role of the non-human takes prominence, which includes the physical setting of the production networks. 

3.       All these networks – including environmentalism, industry, and human habitation – are precarious, fragile, and temporary points of translation, or moments of order that can dissipate from contact with other orders.  The material durability possible with architecture makes it the medium of choice for artifacts intended to weather the dissipation of its production network. 

Blackbox Extensions

One of ANT’s basic concepts is that of the blackbox.  Effectively, a blackbox is an abbreviation of a network system, in which only the inputs and outputs are considered.  The inner workings of the black box are not important to understanding its role within the network it occupies, however, if these inner workings become relevant the blackbox can be “opened” such that the network within it can be exposed for analysis, e.g. if a computer becomes a focus in the inquiry then its blackbox can be opened, and the details of how it produces its outputs mapped, the system can be attributed a design history, compared to other systems, etc. extended to whatever degree needed for the inquiry.  The blackbox is the starting point for the following three extensions of ANT, graphically depicted in the diagram below:

  • Every Actor is a Blackbox.  Every network is composed of actors, but each of these actors is itself a black box that can be opened as needed.  This creates an index of scaling of sorts, in which networks are nested within greater networks, and these within yet greater networks, etc.  There is no theoretical limit to the number of nested steps that can be traced as one extends the network.  However, the fact that every ANT analysis begins at the scale of a human creates an orientation point within the chain structure.  The scale of the human becomes the entry point into the fractal-esque chains of nested networks that extend in both directions: networks that the human in part composes, and networks that compose the human. 

  • The extension of every network approaches nature.  Within the logic of ANT, nature is that blackbox which every instance of network extension inevitably approaches.  All networks, extended far enough, identify dependencies on the presence of oxygen, water, food, resources, etc. across the board -- all the implicit contributions of the surrounding environment, ecosystem, and climate.  Nature is the non-human blackbox whose intermediaries sustain all other networks.  The fact that they routinely fall below the threshold of "negligible effect" on the network only demonstrates their ubiquitous assumption.  For the vast majority of human history, nature could remain a closed blackbox without the possibility of humans altering its outputs; now however the scale of technological production – industrial, chemical, agricultural, etc – is putting stresses on the inner workings of the blackbox such that some of the outputs suggest the need to open it.  Environmental science is the research actor-network whose project is to begin opening this foundational blackbox – whose internal complexity is such that dramatically escalating our technological sophistication, esp. in terms of computation, was a prerequisite for us to even be able to recognize its causal webs.  

  • Inside the black box of human subjectivity is not another network.  A common criticism levelled against ANT by other schools of social theory is its preoccupation with the material details mapped as actors into its networks, the tools, the objects, the environmental features, etc.  However, for the question of a work of architecture, it does not go far enough.  The blackbox of human subjectivity unquestionably contributes outputs that impact networks.  As much as ANT would like to de-centralize subjectivity and maintain the purity of the network model, within these chains of nested networks the human actor is the blackbox that cannot be opened and mapped as another network.  Human subjectivity simply won’t translate to a constellation of actors and intermediaries.  Nonetheless, it is only in the terms of human subjectivity that the force of architecture on those within it can be clarified.  For this, we need an entirely different mode of description: phenomenology.  In order to show architecture’s import to an actor-network that it contains, we must venture outside of ANT and think in the terms of phenomenology.

ANT’s conceptual inclusivity, as applied to an actor-network such as a climate summit, leads to a specific aspect of the non-human within this network that ANT has not yet ever fully addressed.  As will be shown below, especially in the case of any sort of network with a compressed time frame and that brings the human actors physically together face-to-face within the same space as for a global summit, the physical setting itself must be included in the actor-network map due to its influence on the human actors during the process.  From the architectural space itself, to the objects and technology present, environmental conditions, etc. all would require inclusion in the network analysis.  Highlighting the paradox produced by ANT, human subjectivity is the conduit of the intermediary exchange between the architectural/object context and the human network actors -- intermediaries that, as will be shown, cannot be deemed below the threshold of “negligible effect” that would allow omission from the network map.  If properly understood, the physical setting will be mapped as a deeply consequential actor(s) in the network, but to see this we have to open the black box of human subjectivity via phenomenology.

As an interlude, this video was shown at the COP15 Global Climate Summit in
Copenhagen in 2009, in a large domed theatre complete with emotive soundtrack:

Why?  The obvious answer is the correct one: to (re)sensitize the conference attendees to the grandeur of the natural world as they proceeded with a bureaucratic process that could in part determine its fate – something like a heartfelt testimony before jury deliberations.  This simple intention, however, reveals an issue critical to our purpose here.  The official inclusion of this film within the conference programming reveals the implicit recognition of an absence, an absence at the center of one of our most developed, organized, global, and potentially-far-reaching-in-consequence initiatives toward sustainability.