Site Matters

Design Co n cepts,
Histories, and
Strategies

Ed i ted b y
Ca r o l J. Bu r n and Andrea Ka hn

ROUTLEDGE
N EW YO RK AND LONDON
 fining Urban Sites

l11w does an urban site gain design definition? What delineates its
11111daries? How does it engage its surrounds? What determines its
,II\'? This essay works through the problem of site definition as a nec-
11 rily indefinite task, especially when looking at terms of site defini-
1t111 in urban design. As used here, urban site makes double reference
1h11th the whole city and limited sites within it, since even the small-
1 urban design intervention always speaks to the project of city-build-
1111 writ large, and defining applies to both a process a nd its outcomes.
Manhattan Skyline. 1 1 ~sue are means of site definition in urban desig n as well as the site
1111wledge they produce.
In design discourse, the qualifier " urban" attaches to the concept
11 ' ite to no significant effect. This should no t be the case. When rep-
wnting urban sites, or relationships between sites in urban situa tio ns,
h·" gners draw on concepts, terminologies, a nd gra phic conventio ns
li.11 pertain to all kinds of sires, in general. Commo n terms (place,
11111nd, context, scale, location, boundary, etc.) remain largely indis-
11minate with respect to differences in setting or settlement conditions.
Without benefit of language expressly applicable to urban sites, their
•lt'linition, as a subset of sites in general, remains tied to notions of
p1operty and ownership, to a physically delimited and containable par-
11'1 of land. A site is defined as urban adjectivally, based either on geo-
Kr.1phic milieu (an urban design site refers to a limited place within an
ilrcady established urban area or to a n urban area in its entirety) or
11liysical size (urban design sites are presumed to be larger than archi-
11·crural sites and smaller than regional ones).
 R

r \I Ac\ I
'r1••
,........#uu
II 1,., I
Jf,, •...,,,. ""' ''""''• ,I
j}~~
#I lt..7u,bJ.,·• Ji 1•. -.,,
Jtri'/I ,f..,w 1fy'#-

Figure 11.1. Anonymous, Palmanuova Plan, 1713.

To frame a site in explic itly urban terms, 1 use examples fr111 11

York City to lay out an operationally based d efinition concer111 d
what a site " does " in the city rather than what (or where} it "i~." I
I turn co the role of representation in the site d efinition process. I 111
I conclude by offering up new terms co address the complexi1y 111
ent in urban sites. These terms provide conceptual tools appl1: ,1hl
urban analysis as well as urban design. By representing sites n' 111 IIM11rc 11.2. Leonardo da Vinci, View of Milan, sixteenth cenrury.
multiple boundary conditions an d multiple sca les, they franw .1 I
conceptua l mo del for d escribing, interpreting, and analyzing pl lwo drawings, a 1713 anonymous plan of the ideal Rena issance plan
slated for urban design intervention.
111 Palmanuova, and a sixteenth-century Leonardo da Vinci sketch of
~ t ilan, register an often-overlooked but significant distinction in the

DEFINING URBAN SITES

The point is not that drawing boundaries is somehow impern11. 1
ble ... but cha t the permea bility of those boundaries has to bl' lllll
sra n tly reasserted; more than this, that the space in which they ,11
drawn is not a simple pla ne. Each side folds over and implk.111
tlw other in its constitution. 1
"·'Ydesig ners define site limits as well a s how they understand s ite
-1 .ilc. T he Palmanuova plan depicts the urban site as a clearly bounded
pl.H.:e. In this walled enclave intended to be impenetrable co attack, the
111 y is d escribed as a fixed object in a n open fi eld. The drawing's ccn
{l·rcd composition, inset textua l inscriptions, a nd heavy dark li1ws
111dosing fortifications reinforce the reading of a city figure aflo:ll i11
1•111pty space. The plan strongly delineates inside and outside. lll\itk
'11 I
1he walls of this city rendered as discrete object, everything sits c111
ful ly contained in its proper place.
In stark contrast, da Vinci's sketch swirls with the movement ~ 1
many trajectories crisscrossing an unbounded space. Radiating 1111
activate the drawing's surface, projecting an image that extends out
ward beyond the edge of the page. Neither the bird's eye view at th
bottom nor the plan above inscribes full enclosure. This draw111
which depicts a set of active interrelations, makes it impossihl1· I
locate the edge of the city. What lies inside its boundary and whnl II
outside is unclear. The limits of this urban site cannot be pinned tin
in the horizontal or the vertical dimension. Its boundaries r<.:11111
porous, its figure incomplete.
Comparing these historical images illustrates an important difft·1·111
between an idea of site linked to conventional notions of place and 111
disentangled from notions of limited location. The single Palmni11111
plan conceives the city as stable and rigidly bounded. The co1111111, I
sketch of M ilan shows an active setting with permeable limits, an 111 h
site comprised of many overlapping spaces. In Leonardo's image IH I Iii
der divides site from situation. Rather than equating boundary with
line of separation, this sketch encourages viewers to ask how an 111 h
site is linked to its outside. Instead of creating divisions that franu· ~u
ple enclosures, as Palmanuova does, the looser and more porou ~ 11111
of Milan offers an alternative conception of site limits and scale. It 1 I
tures the complexity found in actual urban situations.
Consider, for example, a high-density residential project 0 11 11 I 11
lot extending north from 56th to 72nd Street, on Manhattan's l 1111
West Side. Understood narrowly as legally owned property, thi ~ 1111
developed by Donald Trump-obviously has a fixed boundnl'y lit
H owever, since the urban impact of his development reac h «~ W
beyond the edge of his parcel, when considered in urban 1·~· 1·111 ~, I
significance of this legal perimeter diminishes greatly. Trump I II
urban site includes not only the ground under the residcn1·i11l lrn~ r
but also those areas affected by their construction. For insta1H.'I', 11 111
way station three blocks to the east required renovation to 1111111
modate the expected load of thousands of new co 111n111rcr~, 111111
waste treatment facility eighty blocks to the north in I In rl <:111 w 11 - lit1
to bear the infrastructu ra l burden of Trump's high-d1:nsity 111 \
I k li ni11µ lfrh1111 Sl 1 1·~
forcing changes to New York City's subway and sewage systems, tht·
1woperty limits of the Trump site are hardly impervious to the many
lorces that ultimately establish the project's urban condition. Adopt-
111g an operational definition of the site-based on how it works in,
with, through, and upon its urban situation-alters the understand-
ini; of Trump City's " limits."
Treating urban sites as operational constructs recasts their bound-
t•dness. Instead of demarcating simple metes and bounds, defining
wban site limits requires accounting for co-present, but not necessar-
ily spatially coincident fields of influence and effect. Urban sites encom-
pass proximate as well as nonproxirnate relations, physical as well as
11onphysical attributes.2 As settings for interactions and intersections
lhtlt transgress abstract property divisions, urban sites are conditioned
hy, and contribute to, their surroundings.
Times Square, in New York, easily fits such description: a place
whose identity is comprised by interactions between a global circuit of
rntcrtainment (the Disney Corporation, Conde Nast}, a metropolitan
1 mssroads of commercial developments (Broadway and Seventh
Avenue), and a local district of direct and imaginary engagements
(llroadway shows, Madame Tussaud's Wax Museum, ABC's Good
M1>rning America, the minimal remains of an erstwhile thriving sex
l11dustry). The specificity of this urban site is construed through an
111r:iy of co-present but not coincident operations. Its reality-or, more
t~rn ra tely, its realities-are constituted through the experience of rad-
k 11lly shifting programs in constant interaction. What defines Times
'1111 are as an urban site is a function of the crossings of spatial net-
works, each with its own degree of spatial extension. The determina-
111111 of its boundary-or again, more accurately, its boundaries-
1lt•pc-nds on how far afield these networks, and their influence, reach.
\ ~ .1 11 urban site emplaced in numerous local, global, metropolitan,
1111d n:gional settings, Times Square is tied into diverse scaling processes
11t 1111c time. While it provides a particularly vivid example of the mu!-
11 ~1. d ccl site, urban sites-wherever they are located and whatever their
111• - will be similarly constituted.
I kll 's Kitchen, lying just two blocks to the west, operates at just
1 ~ 11rnny scnlcs. The area is at once a residential neighborhood, a
111111 111<·ri:in l distrk:t, and a nod al intersectio n of t ra nsportation

111lr.1s1n11.:rures. le is the locus of a national hjghway system {enlnll
111idtown Manhattan through the Lincoln Tunnel); regional, w1
rnunrry, and international bus lines {arriving at and departing f11
the Port Authority Terminal); a metropolitan public transit :.pl
{subways and buses); global speculative ventures {proposed l.11
sca le development on the far West Side); citywide commercial 111
kets {specialty food shops, restaur ants catering to immigrant t
drivers); loca l communities {H ell's Kitchen neighborhood, with
own association). Numerous fields of operation converge at tlw. 11
place, each involving different scales of activity. As such, the srnlc
this site cannot be characterized as singly or simply urban. Huth
this place operates at loca l, metropolitan, regional, national , 11
globa l scales. As an urban site it is scaled through a sec of dy1111111
functions created by fluid interactions between many diffen·1111111
extensive processes. 3
Embedded within, and constitutive of, so many framing co 11t1
such multisca lar urban sites open to diverse interpretation.4 Thq
saturated with difference, permeated with irreducible diversity. h
erologica l, to borrow a term from Mikhail Bakhtin.5 They of111
myriad d imensions for consideration (economic, socia l, hi~te111t
physical, political, haptic), each of which situates the site w1ll1111
web of s pecific associations. In terms o f their limits and thci1 ~' 11
urban sites present designers with shifting and potentially co11ll11 ti
identities. As such, they are best characterized as resulting (111111
matrix of forces, impossible to recoup and therefore imp01.,1ltli
resolve.'' 6 They are crisis objects that destabilize our certainty 111 I
real .7
Urban sites are dynamic rather than static, porous rnth1•1 rh
contained, "messy" like da Vinci's Milan sketch rather than " 111 I
like the ideal plan of Pa lmanuova. Defining them in design 11·1·111111h1
does not come down to establishing some unique identity ol 11 1111
ired physical place, but quite the opposite. It involves ret.:og11l:t 1111 ti
overlay and interplay of multiple realities operating at lhl· ~.11111 1111
on the same place. How designers give definition to lhl·•w 1111ilt1
lent and mulriscalar urban design sires, howeve r, rcmnin ' 1111 "I
quest ion.
l>i-111111111 l l1h1111 'lllt- ~II
REPRESENTING URBAN SITES
However forcefully the real and the represented world resis1
fusion, however immutable the presence of that categorical boun<l-
;try line between them, they are nevertheless indissolubly tied up
with each other and find themselves in continual mutual interac-
8
t ion, uninterrupted exchange goes on between them ...
< .1vcn an operational definition of urban sires as multiscalar, heteroglor
l'ltings for interactions and intersections, how do designers think
through their complexity and multivalence? This question raises the
1~,ue of site study and with it the means, methods, and modes of site
ddinition processes. As in any design process, ideas of site come
1hrough making. Designers confront the challenge of de.fining urban
li es through a creative process of representation.
The artifacts of this process, representations such as drawings and
111odcls, do not simply illustrate what designers think about {in this case,
ll1l· city); more profoundly, they reveal how designers think. The iden-
1111t·s of an urban site can be construed many ways. Mappings can pres-
1111 each "reality" separately and attempt to position each in relative
11•1m-; as a function of shared descriptive and analytic parameters {scale,
1l1 .1wing type, categories of information, etc.). Or they can project a
l11·1t·rogeneous urban condition by utilizing representational techillques
il1.1t actively combine distinct parameters. By bringing different realities
11110 contact and establishing methods to chart their interplay, the
p1111..css of site representation works as the staging ground of site think-
111g. Ir is a place of assembly and a point of departure for constructing
11·l.11ions between and across different forms of site knowledge.
111 common usage, representation is a word loaded with meaning; it
111"' politica l, philosophical, symbolic, and aesthetic dimensions, visual
111d 110nvisual connotations. Even in the relatively focused vocabulary
1ii dl·-;ign, representation is a term subject to misunderstanding. Used
'" ,1 11oun, it refers to things made. Used as a verb, it refers to a process
iii 111.1king. Hut rhesc two mea nings still do not make the full extent of
11 p11''l'IH.1t ion's role in site dcfinirion apparent. Representation is a con-
, 1pt11,d iool t h.ll ordtr<; und<:rstandinµ of the mult ival<:nc(' of urbnn

Figure 11.3. Rodolfo Lanciani, Forma Urbis Romae, Plate 15, 1893-190 1
Figure 11 .4. Giambattista Nolli, Rome Plan, 1748.
sites. It is a means of literally thinking through their ma ny l'C11ll11r
presencing as well as positioning them in relative terms.
Site representations propose working hypotheses fo r comp1•1l1111
ing and testing working definitions of urban site. To grasp 1h1 II
import of this idea, one first has to recognize the expa ns ive poll 1111
of representation: that in the most profound sense, reprcscn1 111l1111
not about depicting reality, but about making kno wledge. 1101· d1 ~1.-1
it is a mode of conceptual operation, a process o f k now lcd1w 1111111
tion. More than simply amassing and organizing focrs, liHlll'l'M, 11
impressions of a given condition, the descriptions and n11 11 lyN1'h 1h
JHi)
l111,1g1wrs prod uu• 111 1111111 )' l\1' 111•1111 1' 11 w lrnuwk-dgc lll'Ccssa ry 1o cnp.:1gc
1 wv1·11 co11 di 1ion ll N 11 NIH•. "111• 11•prcsc11ta1ion is not a matte r of gct-
il111: n::1 lity righ1 111> 1t1mh :ii. :1 111,lltcr of constructing forms of knowl-
.1
il ))1· 1hat can cope wi 1h multiple realities. In this sense, site drawings,
111rnkls, a nd discourses are never mere second-order redescriptions of
111111· preexisting condition as much as they are evidence of thought in
t111111arion, a thought a bo ut what the urban site might be.
/\1 the most basic level, representation gives definition to the urban
hi• because it is a process in which different ideas of site settle down
•II 1'Cttle in-perhaps an idea found through urban history, as in
llndolfo Lanciani's Forma Urbis Romae mappings or, perhaps a idea
h 1 t ~l'd o n city form, as shown by Nolli's well-known figure-grounds of
1h1· sa me city. Each mapping proposes an identifiable site reality,
hl'l':lllse each operates as a distinctive mode of site thinking. To ask
hid1 of two representations depicts the real site is meaningless, just
,_ i1 makes no sense to ask which of two ways of thinking is correct.
111111 inet site representations produce different artifacts, but each arti-
1111·1 instantiates a similarly dialogic and creative performance, an
l')( pcriment in contact with the real. " 9 Site representations construct
lit· knowledge; they make site concepts manifest by design.
FIVE CONCEPTS FOR URBAN SITE THINKING
I nr urban design , site concepts matter. More than merely discursive,
1hvy act as powerful tools to structure site thinking. Yet, without lan-
111111gc to discern between different kinds of sites, the ways designers
11•1wcsent and engage with urban sites cannot be situationally derived.
I 1('11cric concepts only allow for generic site thinking. But design dis-
' 1111rse has no specifically urban site concepts on offer. The five new
1111ms o utlined in the following sections conceptualize site~ in mean-
l11glu I ways for urban design.
Mobile Ground
Muc h o f urban design, as a fie ld of design action, involves framing
, 011strul:tive conve rsatio ns among different interests and agents in the
1 It y. 'l'o be effect ive, designers at work on projects with urban aspi-
 /\11d1'l".I " :11111
r:11 ions must a ccount for and negotiate between many playr1
invested in the future of a particular locale: workers, owners, neigh
hors, and builders; politicians, developers, and bankers; preser\ 1
rionists, ecologists, and economises, to name but a few. Each i11111
ested party construes tbe urban design site according to its 0\\ 11
terms, adopting its own preferred modes of represenration.* The)· .111
claim co know it, but one player's knowledge rarely conforms ro 1h
knowledge held by others. Different lenses filter these undersrandi11"
As concerns shift back and forth between various takes on the ~11111
place, these oscillations define a variable field where the constructed 1111J
the real are not opposed. 10 T hey inscribe a mobile ground where mli II
sires arc understood as dynamic and provisional spaces, as poi111~ 1
departure to parrs unknown rather than places of arrival of I 11,,
address. 1·i Conceiving of urban sites as mobile ground foregrounds I h1 I
provisiona l condition, reminding designers that sites remain subj~·u I
change beyond their control. On mobile ground, urban design nlll1111
are best considered in strategic terms-focused on framing urban 11 It
12
tions and structuring urban processes. Mobile ground describes o 'I'·"
of progression, slippage, and continual revaluation, where divcrst• 11 11
ities tip over, into, and out of each other. It is where site boundarit'' 11
site images shift, bend, and flex, depending on who is looking .
Site Reach
The issue of scale is key to the definition of urban sites, influenci1111 lu
designers understand the context of their work and how they dcfi111 11
geographic extent of their areas of concern.** Because urban sitt•\ pl
ticipate in many differently scaled networks at once, talking ;ih11111
urban scale, as a singular measure or the attribute of sonw r11111
* " Muhifamily housing sites are missing from most conccplions of suburbun l11mlM'11111 I
because conventional ways of measuring and understanding urbanized nrcas huvc ob~cm t•d 1h1 11 1I
tiflcation. The high densicies of apartment concentrations relative to surrounding 11rc11~ 111 iii 1 I
houses. for instance, ure not captured by the common mapping tools used by pla1111cr' 1111d 111111!1 11
Ce11sus 1racts, forecast analysis zones (FAZs), and 1.ransportalion analysis 1.oncs ('l'Ai'.•l ~11111 I
geogrnphic units of analysis used for mnppiog-are simply too hirgc IO ca1?lllrc the Np111l11l pllll• 111
of suburban dcvelopmem." P. I-Jess, " Neighborhoods Apart."
0 "The definition of a study area can be seen as a subse1 of 1hc prohlc111' i11v11lw1l l11 11 II
define n problem or formulace a solution to a problem. in geogmphic tcrn1:•. The 111~1dlllli1111 II
establishing the criteria by which the relevnnt geogrophic boundaric' :ire -cl. The 'l'\'111111lh·~ 111 Jh
ally implicit and hidden assumpcions being made ahou1 thc muur'C of 1hc p1llhll·11111111l II• 111111111
to Mich boundnrie~." P. Mareusc. "Study Areas. Site>. :1ml the Gcogn11>h1l' App111m h 11> l'11hlh 1 I
Dc fini11f.\ Urh.111 'i>ln 1111
obscures the multiscalar condition of urban sites. Urban loca les rq~i~
1cr on multiple scalar networks, in some cases at different times, i11
other cases simull'aneously. Site reach measures the extent, range, and
level of interactions between a localized place and its urban surround-
ings. It gauges vicinities of exchange and intersection between places,
reci procal and nonreciprocal relations, inscribed within and contribut-
mg to co-present urban spatial networks. For urban design, the con-
l'cpt of site reach proposes a much-needed alternative to a conventional,
11csted, and hierarchical model of scale that identifies different scales
with diHerently sized territories, and as such obscures the multiscalar
rnndition of urban sires. Urban sites are constructed by a complex over-
1.ty of distinct but interrelated uses, boundaries, forms, and tempora l
'l'quences. In any given locale, variously scaled interactions establish a
1111ique set of linkages to other places. The reaches of a site depend on
1hc spatial and operational extension of t hose associations and con-
Jlt•ctivities that tie it to other places. By situating any limited place
within the space of the city as a whole, site reach reinforces the fact
1h:it any urban design intervention, no matter how limited in physical
~i:ope, participates in a project of city-bui lding writ large.
Site Construction
1\11ho ugh considered a predesign activity, site analysis inevitably pre-
ligures and reflects design intentions.* The logics and values structur-
111~ initial site observations are always and already prescribed by ideas
thout the future modifications imagined for a place, and conversely,
1hr a nalysis process initiates a way of thinking about place that res-
rn1.11cs through all subsequent phases of design.**
I
1 "/\~ 11f1cn as nm. an architect's description of an ex isti ng context will soon underpin a subse-
11111111,cries of decisions co intervene in that context. A ch;iracterizac ion or context smuggles into the
!11 ·IJ' " 1>1occss a sec of contirming values camouflaged as n description of ex isting conditions and
uh'"'""" t:1c1s; 1hc details of any descripcion of context will usually indicate whether the speaker aims
11111•I'""' 111· 1·cjcc1 i1:· S. lscnstadt, "Contested Contexts."
1• "~ i l c analysis. :1t :1 large scale and recorded lhrough detached rational mappings, has given
111 111 " le rc:1di11g' :111d i111crprc1ations drawn rrnm li1'l'l·hond experience nnd from a specific si1e's
111 1.d 1111d l'l'Olol(icul hislnrics. These site-readings form a s1rong conccplual begi nning for 11 design
" l''"'"I'. 11ml 111<' 1c)!blc1c<l l11 1nc111or:1hk drawing' :111tl 111:1ppi11g~ conveying a >il<.:'s physicul pror>-
1111 , 01~·11111011•, 111111'l.'11,11111 iin111c,~io11,," H. Mcyc1 , "Siie ( 'i1111in11,,"
 1\111111 11 I· 111111
111 urb:1 11 design, 111ort· oil t'll 1h:111 1101 , ,\I ll'' .1n· .1l1 ivt•I)' p111d1111•d
co11structio11 is a sire sludy pr0<.:css tha1 yidds ;1 dcsiµ nn l 1111dl'l 'll11111ll
of site through consciously selective viewing. The s i1·t· tkli11i1 ion~ It I
duces are distinct from the design decision rhal n.:sul1s i11 t•s 111llli~hi
project boundaries (the determination delimiting where dcsi1411 111111
physically take place).13 This site study method embraces dcsig11 11111 1111
and asserts the interpretive basis of any site viewing procesi,.
To define a site as urban, the process of site construction .1l u11U
for multiple fields of influence and effect, each with distinct sp111111l Ii
its that in concert construe a territory of design concern. It rt•t't>Hlll
but does not attempt to reconcile, heterogeneous urban orders 1111d I
ics. By not oversimplifying site complexity, this method of si11· ~111
initiates and supports nonreductive urban design actions. Tht•111
analyses underscore the multivalence of urban sites, making of 11 .1 k
iss ue for urban design attention. Regional, metropolitan, and It
architectural; moving or static; large and small scale; close and d1 11
each vantage point brings different aspects of a site to light :111d 1
way of organizing site information (politically, economically, f<11 1111ll
historically, spatially, etc.) results in a distinct site configurario11 . 111 I
vidually, these expose a p redilection toward some combinnrio11 1ii 11
city's myriad characters.* Drawn together, the many approaclw ~ 111
to approximate the multivalence built into the ur ban landscapt:. It uh
tha n conceive of sites as having one single bounding condition, I
construction posits that site boundaries shift in relation to d11 I''
tion-the physical location and ideological stance-of their lwl111hl
It dispels the illusion of the city as either containable or co111ntll11 I
by hypothesizing the urban situation as a p o rous and shifting sp1h 1
Unbound Sites
Any design action for a limited sire in a city is at once inAut•11u d I
and has consequences for, the city as site. The impossibili1 y or 1i.11ltl
* "Pan of preparing a place Lo become a site involves the fonnation of new n11m1111·l••, I 11111111
anyone who observes real esrn1e development is the narrative onslaught Iha! bcgi11~ ahnn'i 11111111.ll 1
as developers and real estate brokers 10111 the benefiis, for example, of 1hcir pmpo><.:cl 111111111111 111 I •oil
ing, its compatibility with urbane lifestyles. and ils prestigious address. Plannct'l< and dt•\l)lt\\'l• llh ,
plici1 in I.his process. Their presence indicates a seriousness of purpose 11nd even in,·vl111hillly Ii• 1111 11
ect. Their reports and images portroy and publicize the new place. l l ic fiN acl ol r\'111\'\11111• 111'11 h•1"
is the narrative remaking of the si1c." R. Beauregard, "From Place ro Sile."
Defining Urban Sites 29 I
11111 oiw 111·h.111 l o~ .dt•, opt·1·.11iona lly, from its surrounds, lends urbnn
l1 .. 1g11 it~ i11hl'l'l'lllly p11hlic dimension, and acknowledging this puhlic
l11111·11sio11 pro111p1s a critical reassessment of how site boundary is typ·
1i .1 11) 1111derstood in design. lrrespective of whether rights to a limited
1
1li•vdoprnent parcel are privately, publicly, or jointly held, design
h 111111s in urba n contexts have consequence beyond narrowly con-
1111l·d limits of legal metes and bounds. The unbound site uncouples
1hl' defi nition of site boundary fro m notions of ownership and prop-
• II ). Ir views site limits as open to configuration according ro various
h11111s and forces of determination.* Rather than drawing a line
lw1 ween urban and site (equating boundary with a line of separation)
111 h:111 designers need to ask how many ways sites are linked to an
outside," to spaces, times, and places beyond their present and imme-
il1.11 c cont rol.** Defining site boundary in terms of a single property
h11r produces a circumscribed figure, contained, isolatable, and con-
11 ullnble (the site de.fi~ed as entity under design control). Designers
llt'l'd instead to recognize border porosities and ro treat scale as a meas-
llll' of boundary permeability. In this sense, the urban site is unbound
hy virtue of its having many different structuring limits simultaneously
111 play, not because its boundaries are simp ly effaced. Urban sites are
11>111prised of multiple fields (areas under design control, areas of influ-
1111:c and areas of effect) each delimited according ro its own opera-
111111al horizons. 15
"Grounds operale with great nuance. They resist hierru·chy. There are no axes. centers or other
11hvi11usly explicit means of providing orientation. Single, uncomplicated meanings are rare. Instead.
1lw1\• are open networks, pnrtial fields, radical repetition, and suggestive fragments that overlap. weave
h•l't'l hcr, and constan1ty trnnsfonn. Within this textural density edges, seams.junctures. and other gaps
11•1·1·:d 111omen1s of fertile discontinuiiy where new relationships migh1 grow. Relationships among
-•111111ds, arc multiple. shifting, and inclusive. They engage the panicular and 1he concrete rather than
1111' .1hs1ract and the general.'' R. Dripps. "Groundwork."
•• ·'Over the pas1 thiny co fifty years. tbeories in the science of ecology have been reconsidered
111 .11 least three major areas: first, wi1h regard 10 whether local ecosystems can be considered "closed"
111 li11 i:cr-scale llows of materials and energy or whether the influences of these larger llows should be
1 "'" idcrcd integral 10 local sys1ems (I will refer 10 this as the spatial scale paradigm shift); second, in
llw 1ki:n:e 10 which local and regional history inOuences contemporary ecosys1em dynamics (i.e., the
11•1111 1111~11 scale shift): 11nd third, in 1he explicit consideration of physical landscape pauerns as an
ll11p1 .. 1an1 comJX>nc111 or ccosyMem functioning (i.e., the pauem shifi). These developments have
l11 11ml i111plica1iu11s !'or l'Ct1lu~i,1 ~ who now think dilTerently about relationships between local obser-
101111111' nntl cvc111s (OI l11rnl •11111111l 11m1ngcmcn1s) and rel:uionships 1ha1ure neither locnl 1101· rccen1."
11 11111 "Shll'll1111 Sl11·~ ..

Urbon Constollotlon
Con text is what tlw si1e is not. Yt't urh.111 'Ill'' t'\l'I .111d lhll I h 11
many contexts.* How, then, 10 dcli nc il1t· i:o11lin1·s ol 11 rh 0111 111111
traditional idea of context implies that sires dcri vl' ddi 11i1i 1111 1111111
larger situation. Seeing a sire " in conrcxr," howcvt·r, dt·iwnd• 1111 I
raining a clear distinction between inside and outside, ilw 1i•li\ 11 1
i~g. the difference between the boundaries of a building 101 11r1 11
limits of an urban site.
At once a concept and a process, urban constellatio11 hlu 1·- 11 11
between context and site by demarcating site interactions n~ 1·111,_ 1
tiple fields of urban operation. It refers to a dynam ic rclat 1011 ii
strucr-formed by myriad interactions between variable (cm.:1·, (pit
cal, social, political, economic, etc.) animated across multiple s~ 11 l1
embedded in local, metropolitan, regional, and global spu11,t1 1
works)-and the process through which that construct is defi11nl f
process of ur:ban constellation involves integrating knowlcdHt' of 11
~lace-b~sed urban cha:acterisrics with knowledge of larger-st.di I
rial logics chat underlie contemporary urbanism in all irs for 11 1
problematizes the received idea of context as some outside, inip11- I
backdrop.
Constructing urban constellations is not simply a matter of 1•11 111
i~g the contextual frame through which a particular place 1111n I
v 1 ew~d. Rather, the concept of urban constellation requires thar d 1 ·~ 1 141
ers situate their urban sires in multiple contextual, or scalar, (1 ,1111
simultaneously. Constellations foreground context itself as a vnr1 1il1I
Further, by projecting site and context as mutually implicated i11 1h
other's constitution, urban constellations reinforce understancli111•\ 1
site as a relational construct. ' 72.
• '.'Few cities or buildings are more rhoroughly documenred than Paris and the work\ "' I
Corbus1~r. Maps n~d ae~al photographs _or .the sccrors of Paris where Le Corbusier's proji•i I ~•
locaied are as read ily ava1laable as the ub1qu1Lous. published versions of 1he building plans, M'l'll"n
a'.1d elevauons. Yer , no_d~un~ntation exisis of 1his architcc1's work as i1 relates to i1s urban sitr 1h
"mple•. y~t huge. o".11ss1on m the otherwise endless sea of infonnarion and spccula1i011 011 I
~·:d1ui..1~r 1~ astomsh1~g. h dem~nsrrates the pernicious obstinacy of a narrow framing of subjc~1 111 ,11
I~ •· '~luch go~s hand 111 hand wuh the modern concept of cmcgorization. Ca1egorizarion lend~ 1111 11
1 111'.'""'~ and 1~olate'. rnther than rela~e. !he 'phenomenon of concordance· referred 1 by Le Cmhll
0 1
"'' "" 111.'lie '.n1ers11ces between bu1ld10g plan and ci1y map. It is here 1hat the story of the ·:icri1111 , 1
'"" wurl. on "'<urround is recorded. and where the 'environment brings its weight to bc•ir" \\
lkdl11'1d, ''The Suppressed Sile.'' ' '
, ..,...,"''' ' " " ....,.. .. ··-
0 ~ININ C4 1111 INOl!FINITE
• 1111trp1' 1111rltt1l·d 111 rlw p1nnl111g wutons to11,idl'r mh.11t dt·-.1gr1
111•1• •1, rd:11ion:1I i:o11s1nri:1s. 111 so doing, rhey oblige rela1io11al ~ i1t·
"1l11l 111g. They invire designers to consider how urban design sites dif
1. 1 1111111 architectural ones on more than simply locational o r dimcn
11111.d ~round s, emphasizing that limited locales in cities incorporate
u1h.111 processes, systems, and logics that qualify and extend to the city
• .1 whole.
111 lieu of adopting topology (or topography) to generate schematic
111· rl·presentations, these new concepts set up a site definition process
101111ded in tropology. 16 Slippages in meanjng between the terms inten-
111111:1 lly figu"re urban sites as dynamic and processive. Their purpose is
11111 10 stabi lize meaning, but to challenge the very idea of a stable
111 h.111 site. These tools frame a new conceptual model for thinking
1ho11t, and thinking through, urban sites. They construe the process of
111han site definition as one of defining the indefinite. Instead of defio-
111g site in a narrow lexical sense, these concepts activate gaps between
1g11 and meaning to characterize urban sites as spatially elastic and
1t·rnporally provisional. Each recasts received ideas of boundary and
-~.tic in a slightly di fferent way, yet all rebow1d around the sa me under-
lying poinr: that for urban design what matters is gaining under-
~1:111ding of the city in the site.
Notes
I. Derek Gregory, Geographical Imaginations (Cambridge: Blackwell, 1994),
l. Melvyn Webber, "Urban Place and Non-Place Urban Realm," in M. Webber
et. a l., Explorations in Urban Structure (Philadelphia: University of Pennsyl-
vania, 1964 ).
I. Neil Brenner has written extensively on the issue of scale and scaling
processes. For a discussion of urban scale, see "The Urban Question as a
Scale Question: Reflections on Henri Lefebvre, Urban Theory a nd the Poli-
tics of Scale," ftlternational j ournal of Urban and Regional Research 24
(2000): 361-378. For a discussion of rhe fluidity of scale, sec Erik Swynge-
douw, " N either Global nor Local: 'Glocal izarioo' and the Poli tics of Scale,"
in K. Cox, ed., Spaces of Globalization (New York: Guildford Press, 1997),
137-166.
4. See Nigel Thrift on the definitio n of context as "active" and "differencially
extensive" in Spatial Formations (London: Sage Publications, 1996), 3.
"'•, /\ nd rca I<:i hn

5. On beterology, see Tzvetan Todorov, Mikhail Bakhtin: The Dialogfr11/ I

ciple (Minneapolis: University of Minnesota, 1998), 56.
6. On heteroglossia, see M. M. Bakhtin, The Dialogic Imaginatio11. 11 1111
Emerson and M. Holquist (Austin: University of Texas Press, 198 l). "'I
7. The notion of the city as a '" crisis-object' which destabilizes our cc11111111
'the real"' comes from Robert Shields, "A Gltide to Urban ReprcNt'lll 11
and What to Do about It: Alternative Traditions of Urban Thco1·y," Ill
King, ed., Re-Presenting the City: Ethnicity, Capital and Culture i11 tl•r
12
Century Metropolis (New York: NYU Press, 1996), 227.
8. M. M. Bakhtin, The Dialogic Imagination, 254. High-Performance Sites
9. Gilles Deleuze, Foucault (London: Athlone, 1988), 12.
10. On the constructed and the real, see Bruno Latour, "Whose CosmoN, WI
Cosmopolitics? Comments on the Peace Terms of Ulrich 111
http:llwww.ensmp.frl-latourlarticles/articlel92-BECK-CK.html (April 11111
11. " Places are best thought of not so much as enduring sites, but as 111w11
of encounter, not so much as presents, fixed in space and time, bu1 11 ~
able events, twists and Auxes of interaction." Ash Amin and Nigt·I I ht
Cities: Re-Imagining the Urban (Oxford: Polity Press, 2000), 30. I'nrol J. Burns
12. See my Aarden Hank, "Nomadic Thoughts 1: LA" LAFORUM N1•11• I
(September 1992): 5; and "Nomadic Thoughts 2: NJ," LA FORUM Nm
ter (February 1993): 7.
13. On determi.ning urban design site boundaries, see Edmund Bacon el 111.. I DYNAMIC ROLE OF THE SITE IN BUILDING
City Image," in Elizabeth Geen, Jeanne R. Lowe, and Kenneth Wnlk1•1, 1
Man and the Modern City (Pittsburgh: University of Pittsburgh P r·c~N, 111 Nn1 fixed or static, the role of tbe site in the design disciplines has
14. For a more extensive discussion of site construction, see my " 1•1'11111 I .t lt ered over time. This essay, an exploration of the changing status of
Ground Up: Programming the Urban Site" in Harvard Architectural U 11
~ ltt· in design and construction of buildings, assumes the site to be both
10 (1998): 54- 71.
15. For a more elaborate discussion of the differem areas of a design ~ II• llhtlcrial reality and a cultural construct. As a material reality, the site
the editors' introduction, "Why Site Matters." 1•111lu res in lowly earthiness, but as a living form it regenerates con-
16. On the theory of tropes, see Hayden White, Metahistory: T he 111•11•1 I l1111ously in triumphant potency. As a construct of culture, the site is
Imagination in Nineteenth-Century Europe (Baltimore: Johns Hopkl11- I
versity Press, 1973), 31-38. 111111prehensible only insofar as it is touched by human practices. We
I 11ow the site because we make and shape it, socially, economically,
111d po litically. Both of these viewpoints, realist and idealist, inform
111y .1 rgument, which traces three fundamentally dissimilar renderings
111 ,itc in a rchitecture. The first characterization of site is that of source
lt11 bui lding materials and energy for construction. T he second depic-
111111 is chat of repository for building materials and energy imported
11111 11 for a field. Finally, drawing on an emerging environmental impe-

111'•, I he site is portrayed as linked interdependent systems combining

11111 i11sic a nd extrinsic resources. These versions of site have risen to
l111porrn 11cc, rcspccl"ivcl y, in three different eras: the fi rst beginning in
p11·hi:-.tory and continuing toda y; the second beginning with ind11stri-
dl1,1Jlon nnd rnod1.: rnis111 ; :ind the rhird beginning in the prcscnr 1·inw,