179

FROM SITE TO TERRITORY

LOLA SHEPPARD
Recent discourse and practice in architecture has increasingly proposal for a new understanding of site is provocative, it
sought to ally itself with the fields of urbanism, infrastruc- remains ultimately rooted in a legacy of the genius loci —–
ture, landscape, and ecology. Yet the space in which each teasing out the inherent, but unknowable qualities of a site,
of these disciplines traditionally operates varies in scale, understood as an exclusively physical host.
jurisdiction, complexity, and constitution, therefore complicating Given the complexity of contemporary ecological
architecture’s agency in such a relationship. Typically, and urban systems, and the increasing marginalization of
architecture has been consigned to site; urbanism and architecture within an expanded discourse regarding spatial
landscape architecture are defined by context; and infrastruc- networks, this earlier conceptualization of architecture’s
ture and ecology operate in the territory. If architecture is to relationship to site is, twenty years onward, too limited in
engage other spatial disciplines in contemporary terms, it scope. As we accept the notion that we have entered an
will increasingly need share their protocols; to be resilient, Anthropocene era, one in which no region of the globe is left
adaptable and responsive to changing environmental condi- untouched—–whether directly or indirectly—–by humankind’s
tions. Architecture can no longer define its parameters and impact,3 every site must be understood as the palimpsest of
responsiveness at the scale of its immediate site, but rather, forces. While systems such as ecology, mobility and technol-
must operate at the scale of the broader territory, a space ogy manifest themselves in spatially tangible ways, others,
expanded and thickened with environmental data, competing such as political and economic forces, operate often unseen
social and political claims, economic forces, systems of and by spatially remote means. Each, however, is central in
mobility, ecological systems, and urban metabolisms. This shaping aspects of the built environment, whether financing,
new territory is expanded as each of these influences brings site selection, program, or construction methods. Traditional
its own spatialization, producing composite, fluid boundaries; conceptions of site do not explicitly acknowledge the more
and, it is thickened by strata of information, which extend intricate and often invisible forces at work on the environ-
from the subsurface depths of geology to aerial resources. ment. Territory has become the necessary scale required to
In this scenario, territory is understood as a series of nested register and engage the complexity of networks and informa-
scales from the local, to the regional to the global. tion at play in a given physical environment. For architecture
to think at the scale of territory does not require an amplifica-
After Site tion in size, but rather, a conceptual shift; it demands that
In her essay “On Site”, from the 1991 publication Drawing, architecture, regardless of its actual scale or extents, engage
Building, Text, Carol Burns untangles recent understandings its extrinsic environment.
of site. She suggests two paradigms for apprehending site: It is useful to acknowledge the root meanings of both
the “cleared” and the “constructed.” The “cleared site” is site and territory to attempt to arrive at new definitions or
based on the assumption that site is neutral ground, devoid interpretations. Site, which comes from the root of sinere
of content or meaning, and dependent upon a planimetrically- meaning “let, leave alone or permit,” is defined as “place or
based mathematicization of land. The “constructed site,” position occupied by something.”4 Territory, on the other hand,
Burns argues, emphasizes the visible physicality, morpho- has more contested roots. Terra means land, earth, nourish-
logical qualities, and existing conditions of land, looking at ment, sustenance; but the root verb terrere means to frighten,
the natural and human forces that have shaped the land. In to terrorize. Territory is land occupied by violence.5 In French,
emphasizing the visual however, what is not directly identifi- there exists the distinction of the terroir, and a territoire. A
able is ignored. terroir designates a geographic area homogenous in terms
Burns tracks the multiple notions and terms historically resources and production—–often related to agricultural pro-
associated with site—–the lot, the plot, location, position—–all duction—–and tied to distinct cultural communities. A territoire,
spatially and visually-based, and quantitatively determined. on the other hand, is understood in two categories: a political or
Burns posits a ‘third’ way, which proposes to look at the jurisdictional region, often defined by a border, and natural
specificity of each site, “to discover its latent qualities or poten- territories, defined by more ambiguous ecological delimitations.
tial.” 1 In this situation, an intervention becomes a reading, The notion of territory becomes more complex when
dialogue and critique of existing conditions, and concludes considering the condition of globalization—–the spaces of
that site is “a work, a human or social trace.” 2 While Burns’ information exchange, of international trade and retail, or

1. Carol Burns, “On Site: Architectural era, see Jan Zalasiewicz, Mark Williams, “Are 4. Etymology of site, http://www.etymonline.
Pre-occupations,” in Drawing, Building, Text, we now living in the Anthropocene?” GSA com/index.php?allowed_in_frame=0&search=
ed. Andrea Kahn (NY: Princeton Architectural Today, v.18, no.2 (Feb 2008), http://www. site&searchmode=none (accessed September
Press 1991), 155. see.ed.ac.uk/~shs/Climate%20change/Geo- 15, 2011).
2. Ibid.,164. politics/Anthropocene%202.pdf 5. David Delaney, Territory: A Short Introduction
3. For further discussion of the Anthropocene (accessed September 20, 2011). (Malden, MA: Blackwell Publishing, 2005),14.

GOES SOFT bracket

78 SHEPPARD PORTFOLIO
 180 formatting/distributing

of military occupation. How, for instance, can we define the re-scripted to construct new understandings of the territory,
site of Walmart, a military base, or a displacement camp? forming a feedback mechanism of information collection
In these scenarios, individual sites or nodes are so inte- and production.
grated and dependent on larger economic or technological Architect and inventor Buckminster Fuller understood
networks, or the by-product of geopolitical negotiations, the capacity of maps to oscillate and shape our reading of
that node and network merge into ambiguously delimited the territory. In his essay “Fluid Geographies” from Ideas and
boundaries; sub-sites are nested within sites, operating at Integrities (1963), Fuller sets up the dichotomy of sea and
multiple scales. One might equally ask, what are the limits land to describe the dymaxion map he first published in 1927.
of a site when engaging questions of ecology? What are The dymaxion representational system was the substrate by
a site’s boundaries with respect to hydrology, toxicity, or which to understand territorial relations and networks—–energy,
species migration? Ecosystems are increasingly understood resources, migration patterns—–anew. He presented two versions
as dynamic, subject to the influence of larger-scale flows of of the map: one that prioritized the relationships of land masses
materials and energies.6 Others envision ecosystems as as a nearly continuous chain (in which all projective error is
sequences of scalar relationships, where sites display variable massaged into the oceans); and the other prioritizing the
levels of stability and dynamism according to their size.7 oceans as continuous. Fuller’s recognition that our reading of
territory can oscillate is fruitful in embracing a more complex,
New Cartographies: Representing Territory “softer” reading of the term. The implication of Fuller’s maps
Given the extent and often fluid nature of territories, cartographic is that geography can be understood as dynamic, and that the
representation becomes a primary method for apprehending cartographic act should incite new modalities for documenting
the complexity and mutability of forces at work, but equally, in and apprehending complex territories.
projecting potential futures. In some sense, the cartographic Embracing the need for an expanded and more fluid
act becomes both the synthesis of information and the initial understanding of context, and as a counterpart to Burns’
registration of a potential (territorial) site. Not surprisingly, distinctions of site twenty year ago, I would offer two
Burns’ understanding of site also depends upon its repre- conceptions of territory that have driven our contemporary
sentation: the planimetrically-driven cleared site versus understanding of context—–the “layered territory” and the
the sectionally-understood constructed site. Implicitly and “networked territory.” These two models correspond both to
explicitly, Burns acknowledges: “landscape and survey inform methods of analyzing and representing, but equally, to ways
ways of seeing because they are forms of knowledge. Like of conceptualizing territories.
architecture, they frame information or content; they control by
establishing principles that make the world comprehensible.” 8 Layered Territory
In the ensuing twenty years, our capacity to apprehend The layered territory stratifies its environment into a series of
the complexity of territory has been amplified by an ever- individuated layers and systems. It enables an in-depth exami-
increasing amount of data available. Access to environmental nation of the systems—–physical and natural—–at work within
reports, ecological studies, economic analyses, and seasonal a site, but with little focus on the interconnection or potential
data, for example, offers an opportunity on the architect to overlaps of these systems. Landscape architect and planner
acquire this knowledge, but also the need to sort, arrange, Ian McHarg brought this methodology to its apotheosis in his

WRITING
and prioritize this information in order to then find productive seminal book, Design with Nature (1969), which provided
points of intersection. strategies for disaggregating a territory into its multiple uses.
Anthropologist Gregory Bateson, from his essay “Form, Such methodology focused primarily on patterns of land
Substance and Difference” in Steps to an Ecology of Mind use and human settlements, and formed a precursor to the
(1972), elucidates the essential impossibility of knowing what Geographic Information Systems (GIS), which decades later
constitutes the territory, given that any understanding of it is remains a powerful tool for the analysis of layers of informa-
based on some form of representation. Bateson posits that tion that can be spatially cross-referenced.
“the map is different from the territory. But what is the terri- McHarg is widely credited with a revolution in the field
tory? What is on the paper map is a representation; and as of landscape architecture, shifting it from garden design to
you push back, what you find is an infinite regress, an infinite multi-disciplinary territorial design engaging ecology, geology,
series of maps. The territory never gets in at all.” 9 Bateson and anthropology. However, McHarg’s stance throughout
points out that the usefulness of a map (a representation of Design with Nature is largely anti-urban. While recognizing
reality) is not necessarily a matter of its literal truthfulness, the inevitability of the built environment, his interest is in
but in its having a structure analogous to the territory.10 The restoring the centrality of nature and ecology as the systems
cartographic act, in this sense, becomes the code-work for that determine territorial organization.
comprehending the territory. However the code-work can be

6. For a further discussion of site ecology 7. For a more detailed discussion of the rela- 8. Burns, Drawing, Building, Text, 161.
as set of dynamic processes, see Kristina tionship of time, space, scale and ecological 9. Gregory Bateson, “Form, Substance and
Hill, “Shifting Sites,” Site Matters: Design stability, see Richard T. T. Forman, Land Difference,” Steps to an Ecology of Mind
Concepts, Histories, and Strategies. eds. Carol Mosaics; The Ecology of Landscapes and (Chicago: University of Chicago Press, 1972),
J. Burns and Andrea Kahn (New York: Regions (Cambridge: Cambridge University 460.
Routledge, 2005), 131-152. Press, 2005), 7-14. 10. Ibid., 460.

Sheppard From Site to Territory

SHEPPARD PORTFOLIO 79
 181

Staten Island Study: Existing Land Use; Existing Vegetation; Tidal Inundation. (Image courtesy of Ian L. McHarg Collection,
The Architectural Archives, University of Pennsylvania)

Potomac River Basin Valley Study. (Image courtesy of Ian L. McHarg Collection, The Architectural Archives, University of Pennsylvania)

In his 1966 proposal for Staten Island, as in many urban and landscape projects today. Often, however, these
of his other studies, McHarg documents—–through maps, analytic diagrams read as due diligence rather than a produc-
charts, and geological sections—–climate, physiography, soil, tive strategy for the reconceptualization of a site. To understand
geology, hydrology, vegetation, wildlife, and land uses. Within the land as a series of separate strata suggests the possibility
categories, gradients are further articulated, from most to that such forces can be understood as operating independently,
least suitable use, in order to prioritize with more refinement when in fact, they exist in mutually interdependent states.
the location of development and nature preservation. The McHarg’s most projective drawings, in a sense, are annotated
maps expose intricate calibrations of uses, which result from perspectives and geological sections, in which the layers of
“asking the land to display discrete attributes which, when the landscape are resynthesized into more complex drawings.
superimposed, reveal great complexity.” 11 Design, in this They suggest tools and possibilities for conceiving of overlaps
scenario, is determined by a system of prioritization rather and transformation of the original layers.
than negotiation. Systems are allowed to coexist, but never Our built and natural environments have become
does McHarg envision overtly hybrid conditions. increasingly intertwined since McHarg authored Design with
The legacy of McHarg’s approach is evident in the domi- Nature. Urbanization in various mutant forms has spread
nance of his representational system in many large-scale algorithmically across the territory, leaving very little of the

11. Ian McHarg, Design with Nature (Garden

City, NY: Doubleday/Natural History Press,
1969), 115.

GOES SOFT bracket

80 SHEPPARD PORTFOLIO
 182 formatting/distributing

Sassen, and Michael Hardt and Antonio Negri, amongst

others, describe the physical and social implications of global
economic and political networks. On the other hand, under-
standing territory as part of complex terrestrial ecosystems
might be understood to produce a reterritorilization—–an
anchoring of territory into locally specific, if dynamic, physical
and ecological networks.13
Benton Mackaye, an early twentieth-century American
regional planner, was an early advocate for comprehensive
thinking about territorial networks and what he termed “liquid
planning”—–an organizational armature capable of negotiat-
ing and accommodating infrastructures and environments
in flux. Liquid planning consisted of two essential spatial
protocols: “watersheds” and “waysides.” Watersheds were
territories whose boundaries could be elastic or porous.
Waysides were large strips of public land running adjacent
to highways, conceived of as “critical conduit between towns
and wilderness or recreation.”14 MacKaye understood these
protocols as ways to structure not just land and water bodies,
but also electricity, mobility, distribution of urban centres and
population migrations, allowing for the organization of large
territories in more fluid ways. It also enabled him to envision
Plan for Appalachian Trail, organizing mobility systems, the built and natural environment at multiple, nested scales,
waterways and metropolitan centre. Image courtesy of Dartmouth calling for global environmental organizations, continental
College Library. treaties, and regional networks.
MacKaye is perhaps best known for the conceptualiza-
tion and design of the Appalachian Trail, developed during
natural world untouched. Increasingly, the need to recognize the 1910s through 1920s, as a terrestrial infrastructure
and conceive of hybrid systems seems essential. What might network that sustained ecological, cultural and economic
a layer merging infrastructure and ecology resemble? To ambitions. MacKaye’s plan consisted of a footpath run-
which layer does one allocate green roofs, or mechanized ning along the entire length of the mountain ridge crest,
agricultural, or water networks which are both manipula- which connected to rail and road networks, linking cities
tions of hydrology and infrastructure? Current GIS software and economies. The proposal, which placed the footpath
equally encourages a model of layering and isolating distinct as the spine of the system, represented an inversion of
conditions. Perhaps the future of GIS lies in users being traditional planning hierarchies that would typically place
able to join categories of information, to produce new layers, rail and highways as the primary order of mobility and trails
new terminology, and new spatial formats. For it is in these as secondary. The territory is understood and organized as a

WRITING
synthesized layers that new urban forms, building typolo- reservoir of natural resources—–recreation, minerals, forests
gies, and landscapes organizations will materialize that are and hydro-electricity that need to be managed and orga-
capable of responding to today’s complex physical and nized. MacKaye proposed “no master plan, but rather an
environmental conditions. ordering principle for a new economy… a new mechanism
for the migration of populations, as well as the economies
Networked Territory of production and distribution.”15
Networked territories, in counterpoint to layered territories, MacKaye described his field of work as geotechnics, or
conceive of context as the subject and product of multiple “the applied science of making the earth more habitable,”16
intersecting networks of human and natural ecologies. The with respect to physical, economic and social systems. This
networked territory understands sites as hubs or pieces within new science suggests that the land, at a territorial and often
much larger territories—–the juncture of multiple forces—–poli- continental scale, had to be planned, and it acknowledged
tical, economic, social and ecological—–whose source and that the design of territory extended beyond the dialectic of
destination are often far from the site itself. On the one hand, natural and man-made spaces. Although MacKaye envisioned
the hyper-networked territory produces a deterritorialization; complex relationships of territorial networks, it is unclear how
a fluid space in which globalizing forces move, independent these might have manifested themselves at a more local-
of cultural or political borders.12 Manuel Castells, Saskia ized scale. For instance, MacKaye’s proposal for “townless

12. For a more extended discussion of empire 13. See Hill, Site Matters and Richard T. T. 15. Ibid., 25.
as a de-territorializing force, see Antonio Forman, Land Mosaics. 16. Benton MacKaye, From Geography to
Hardt and Michael Negri, Empire (Cambridge: 14. Keller Easterling, Organizational Space: Geotechnics, ed. Paul T. Bryant (Urbana:
Harvard University Press: 2000), xi-xvii. Landscapes, Highways and Houses in America University of Illinois Press, 1968), 110-111.
(Cambridge: MIT Press, 1999), 57.

Sheppard From Site to Territory

SHEPPARD PORTFOLIO 81
 183

The network of anchors below the “landscape” suggest the

forces at work on a site—–not producing literal topological
manipulations, but rather conceptual transformations of
use, occupancy, and operations. It is worth noting that the
anchors in Waddington’s diagram are tethered no only to
the surface above, but to each other, suggesting a territory
thickened by forces, which are dynamic, responsive, and net-
worked to each other. Any change in parameters is relayed
throughout the system and in turn recalibrates the surface.
The “epigenetic territory” is one in which forces, data and
information are continuously scripting its performance. How
that information is understood and how it informs a design
response offers a way of thinking about architecture operating
in an expanded territory.
The question then emerges, do there exist models for
such practice? The Dutch firm MVRDV has defined their
career leveraging the role of data and information—–visu-
alized and spatialized—–as a tool for urban and territorial
planning. They have been unique in recognizing the potential
design agency of the complex forces at work in a terri-
tory—–economics, population distributions, zoning, land use,
infrastructural networks, etc. Information directly leverages
program and becomes site. MVRDV represent perhaps the
first generation of architects attempting to give form, though
somewhat literally, to the epigenetic territory.
In their sequence of projects that form part of NL Stadt
(1997-2004), MVRDV analyze the Netherlands across a
host of criteria, acknowledging the country’s tendency to
organize and manipulate its territory. In the face of increasing
bureaucracy, they propose to streamline hierarchies and con-
Epigenetic Landscape, seen from above and below. Conrad ceive of the entire country as a single city or territory. They
Waddington, 1957. (Image courtesy of Edinburgh University divide and analyze the territory into a coast zone, energy
Libraries, Special Collections Department) zone, forest zone, airport zone, port zone and so forth, and
propose a series of urban speculations in relation to each
of these programs. While they engage a range of complex
highways” and “highwayless towns” deliberately separated forces in the overall NL Stadt project—–they isolate each
mobility networks, settlements and green spaces, refuting the force or layer of information to form the basis of indepen-
idea that systems might intersect or cross-breed, as his large- dent investigations, both for testing purposes but equally, to
scale visions and rhetoric suggest. produce more hyperbolic proposals. One of many investiga-
tions in NL Stadt includes Glass or Greenhouse Zone. Here,
Epigenetic Territory MVRDV test new distributions for Holland’s notorious fields
If the layered territory acknowledges the vertical stratification of greenhouses, and in the Zuidplaspolder, propose their
of sites, and the networked territory envisions a horizontal redistribution into ribbons along highways, freeing up land,
intersection of systems and boundaries, the question emerges: serving as noise buffers to housing and capitalizing on less
does there exist a conceptual model that envisions a three- desirable real estate. NL Stadt consists of dozens of such
dimensional field of forces feeding back on each other? And speculations, which together form a composite vision for
if so, what might be the spatial implications? alternative models of urban development in Holland. Despite,
Sanford Kwinter, in his 1993 essay “Soft Systems,” or perhaps because of the unrelenting rigour of approach,
co-opts biologist Conrad Waddinton’s model of “epigenesis” MVRDV’s design system results in a rather literal use of
to describe how differentiated form emerges from a formless data to produce urban patterns and architectural form, yet
or homogenous environment. Waddington’s diagram of the MVRDV openly recognize such projects are speculations and
epigenetic surface attempts to visualize how genes act as provocations. These datascapes suggest that traditional archi-
anchors or triggers in a complex, dynamic system of non-linear tectural design plays a very limited role; it is society in all its
interrelationships. The ball or blastula (the sphere of cells complexities that shapes the territory in the most detailed
formed during the early stage of embryonic development), way. Less design than the spatialization of data, MVRDV’s
is kept moving along the topological surface, in a general speculative work probes the limit states of the forces acting
trajectory of evolution, but one continuously recalibrating to on the epigenetic territory.
external forces. A young generation of architects and landscape
Waddington’s epigenetic landscape offers a potentially architects are testing a more evolved understanding of the
fruitful diagram by which to imagine a new model of territory. epigenetic territory, as evidenced in the pages of this issue of

GOES SOFT bracket

82 SHEPPARD PORTFOLIO
 184 formatting/distributing

Glasszone, reorganizing the greenhouses of the region into linear corridors. Image courtesy of MVRDV.

Bracket. From reformatting urban snow collection and river

dredge material, to the harvesting of wind power and the
processing of ocean water, from the subversion of national
sovereignty through soft landscape strategies to the co-
option of the military as ecological agent, the question of the
architect’s role within a site, a boundary, and a territory are
provocatively tested in this edition of Bracket. In each case
design confronts a seemingly impassable constraint—–geo-
political limits, air rights, ecological thresholds—–and it is in
these moments that the architect becomes active agent not
only in operating in the epigenetic territory, but in designing
it; extending its boundaries and performance. To represent
the territory is to understand it, is to operate within it, is to
(re)design it.
If architecture is to engage the spatial and operational
complexities of urbanism, infrastructure, landscape, ecology,

WRITING
and industry, amongst others, it must take on an expanded
role, capable of reading the composite territories of a project,
negotiating contingencies, and scripting alternative outcomes.
The architect in this scenario is less form-maker than spatial
‘programmer’ capable of negotiating the thickened territory
of a project. Architecture must envision spatial formats and
processes capable of relinquishing control and responding to
their wider environment. Then, architecture will be able to
reorganize the epigenetic territory, acting as the registration
and interface with larger forces, with the intent of producing
new spatial systems, building typologies, and public realms.

Y R OT I R R E T OT E T I S M O R F
Sheppard From Site to Territory

SHEPPARD PORTFOLIO 83