5 dovey and dickson Journal of Architectural Education,pp. 5–13 © 2002 ACSA, Inc.

KIM DOVEY AND SCOTT DICKSONUniversity of Melbourne

Architecture and Freedom?Programmatic Innovationin the Work of Koolhaas/OMA

The work of Rem Koolhaas/OMA has long been identi�ed with the attempt to break with architec-tural ideologies embodied in spatial programs. Programmatic innovations include the production of�elds of social encounter, new functional juxtapositions, and forms of spatial segmentation. Theseare designed to resist the role of architecture in reproducing social roles and structures—to enablecertain freedoms. This paper is a spatial analysis of two recent projects that reveal both achievementsand limits to this project. Koolhaas’s work succeeds through a certain magic; and some of the free-doms are illusory.

IntroductionKoolhaas’ designs are blatantly straightfor-ward. . . . [O]ne and only one cultural aimdrives the work . . . to discover what real,instrumental collaboration can be effectedbetween architecture and freedom.

Kipnis1

The success of the work of Koolhaas/OMA inprofessional discourse rests strongly on the claim tobe an architecture of emancipation.2 This paper isan examination of this claim and a critique of twoof his buildings through the lens of an adaptedmethod of spatial syntax analysis. This may seem anodd coupling because spatial syntax analysis, asdeveloped by Hillier and Hanson, is a structuralistcritique of buildings and urban structures thatwould surely be anathema to Koolhaas.3 Yet ouradaptation of such methods is not intended to bereductionist. Rather, it is an interrogation and exca-vation of the spatial program and a critique of theways in which architectural ideologies may havebeen deconstructed, reconstructed, or reproduced.To do this, we will �rst set aside any formal critiqueof Koolhaas’s work except as it informs this task,not because such formal expression is any less inter-esting or innovative, but because it is bettertheorized. Indeed, it is all that many critics see ofhis work.

Buildings as Spatial “Fields”The works of OMA have been termed the “socialcondensers of our time.”4 This re�ects a return tothe early modernist imperative toward an architec-ture that would remake the habitat and habitus ofeveryday life. However, this is not a return to thesocial engineering re�ected in ideas like the “socialcondenser.” Rather, it is a vision fueled by theformal and social multiplicities of urban life, a visionre�ected in the name of the Of�ce for MetropolitanArchitecture, which we read as both an architectureof the metropolis and an insertion of the metropolisinto the architecture. Koolhaas’s work is stronglyordered by trajectories of movement through thebuilding. The role of vertical movement via escala-tors, stairs, ramps, and lifts is a key to the orderthat is set up, as they become the modes of accessto �elds of encounter or “event-�elds.” Koolhaas isinspired by the notion of an architecture of libera-tion in terms of the multiple “freedoms” for newforms of action that architecture is seen to makepossible. Space is programmed for inde�nite func-tion and chance encounter. Koolhaas seeks anarchitecture that can resist the imperative tobecome a diagram of social and institutional struc-ture, which he terms social mimesis. For Kipnis,Koolhaas’s version of “freedom” is not an overtresistance to authority but rather a form ofprogrammatic sabotage:

More like a sadist than a surgeon, he hasbegun to knife the brief, hacking away its fat,even its �esh, until he has exposed itsnerve. . . . The focus on these reductions isalways on disestablishment, that is, always onexcising the residues in the project of un-warranted authority, unnecessary governanceand tired convention. Reductive disestablish-ment provides the crucial stratagem in each ofKoolhaas’ recent projects, the intellectualmodus operandi by which the architect beginsto transform the design into an instrument offreedom.5

Koolhaas seeks an architecture that encourages anirruption of events, social encounters, and opportu-nities for action. Rather than designing with aparticular hierarchy of spaces and narratives ofspatial movement in mind, he generally workstowards a spatial structure that allows a multiplicityof choices for pedestrian �ow and encounter. Kool-haas wants to “liquefy rigid programming intonon-speci�c �ows and events . . . to weave togetherexterior, interior, vestigial and primary spaces into afrank differential matrix that rids the building of thehackneyed bourgeois niceties of cosmetic hierar-chies.”6

Koolhaas often designs interiors as if they wereexteriors, importing lessons from exterior urban

Architecture and Freedom?Programmatic Innovation in the Work of Koolhaas/OMA

6

space into interior space. These interiors are oftendesigned as �elds of play or arti�cial landscapesthat dissolve boundaries between inside andoutside, between architecture and metropolis. Suchspaces are often functionally open and visuallytransparent to maximize social encounter. Jamesonsituates Koolhaas’s work in the context of theprevailing social dialectics of publicity/privacy andfreedom/ control in what he terms the post-civilsociety. He suggests that Koolhaas’s work enablespatterns of free play within a rigid spatial order:

the originality of Koolhaas is that his workdoes not simply glorify differentiation in theconventional pluralist ideological way: ratherhe insists on the relationship between thisrandomness and freedom and the presence ofsome rigid, inhuman, non-differentiated formthat enables the differentiation of what goeson around it.7

There is an interesting connection here with whatAllen suggests is a shift in architectural thinkingfrom a focus on the architectural object to a focuson �eld relations, paralleling the development of�eld theory in mathematics.8 A �eld consists ofcontingent relations, forces, trajectories, andpatterns of movement such as those that govern a�ock of birds. Allen describes �eld conditions as“any formal or spatial matrix capable of unifyingdiverse elements while respecting the identity ofeach. . . . Field conditions are bottom-upphenomena: de�ned not by overarching geometricalschema but by intricate local connections.”9 The�eld is a material condition rather than a discursivepractice. Allen draws analogies between �eld theoryand architectural attempts to encourage a sponta-neity of action. He suggests that systems with“permeable boundaries, �exible internal relation-ships, multiple pathways and �uid hierarchies” arecapable of responding to emerging complexities ofnew urban contexts.10 A major innovation in Kool-haas’s work lies in the extent to which he has

utilized such strategies in the interiors of buildingswhere they contribute towards the emergence ofnew kinds of social space. The promise here is thatthe �eld-like nature of Koolhaas’s work opens upthe work to multiplicities of experience and action.This idea of the building as a “�eld” rather than anarchitectural object entails a shift in critique fromform to spatial analysis. To what extent do thesedesigns restructure social space or reproducefamiliar spatial structures?

Spatial SyntaxMethods of spatial syntax analysis, �rst developedby Hillier and Hanson, represent an attempt toreveal a deep social structuring of architecturalspace.11 From this view, buildings operate to consti-tute social organizations as spatial dispositions;architecture mediates social reproduction throughspatial “genotypes.” These are not formal types orarchetypes but clusters of spatial segments struc-tured in certain formations with syntactic rules ofsequence and adjacency. Genotypes are seen asinstitutionally and epistemologically embedded. Theforms of schools, of�ces and houses are reproducedfrom a limited number of spatial genotypes. Each ofthese is linked to speci�c social institutions (school,corporation, family) embodying forms of knowledge,production, and reproduction. The work of Hillierand Hanson is widely perceived within the �eld ofarchitecture as positivist and reductionist. Althoughwe share some of these concerns, Koolhaas’sprogrammatic innovations demand analysis and

critique in terms of the link between spatial struc-ture and institutional authority. Syntactic analysis isthe most sophisticated available.

Not enough space is available here for morethan a cursory account of spatial syntax, and thereare a range of analysis techniques. Our analysis is aloose adaptation of syntactic analysis that translatesthe building plan into a diagram of how life andsocial encounter is framed within it.12 Figure 1shows how similar plans with different access pointsyield quite different syntactic structures and illus-trates three primary cluster relations: the string (oren�lade) with no choice of pathway, the fan (orbranching) structure with access controlled from asingle segment, and the ringy network or permeablestructure with multiple choices of pathway. Archi-tecture inevitably involves combinations of thesethree. The linear syntax is an en�lade of spaces withcontrolled movement, which is common in tradi-tional centers of power (such as Versailles) and insome modern retail buildings (with an entry at oneend and an exit at the other).13 The network syntaxis de�ned by a choice of pathways and is oftencalled permeability. The fan is characteristic ofbureaucratic organizations with large numbers ofcells controlled by a hallway. When a linear syntax iscombined with a fan, the result is a tree, a linearseries of fans. A key dimension of syntactic analysisis the degree of network connectivity or “ringiness”versus a tree-like hierarchy of spatial control. Thenetwork structure is de�ned by its multiple andlateral connections, many possible pathways

1. Spatial Analysis Lines, Fans, and Networks

7 dovey and dickson

through it, and dispersed control. Tree-like struc-tures control circulation and social interaction incertain key access spaces. Thus, a hallway or foyerthat is the only access to a cluster of rooms has ahigh level of control over the �ow of everyday life.The permeable network or ringy structure offersmany possible pathways and diverse encounters; the�ow of life through space is only loosely controlled.

Another key characteristic is the depth or shal-lowness of any segment from the nearest externalentry points and the overall depth of the structure.A deep structure requires the traversing of manysegments with many boundaries and points ofcontrol. The diagrammatic method shows the spatialsegments of the building layered into levels ofdepth so that the level of a space indicates theshortest route from the exterior. Depth is an impor-tant mediator of social relations both betweeninhabitants (kinship relations or organizational hier-archies) and between inhabitants and visitors.Domestic space is often structured along age(adult/child) and gender divisions in its deepersegments, while mediating contact between insidersand visitors in shallower space. The syntax of disci-plinary institutions (prison, hospital, asylum, school,and factory) locates subjects under surveillancedeep within the structure.14

Many contemporary buildings, those of Kool-haas among them, are designed with �owing andfragmented spaces, pursuing deliberate ambiguitiesof enclosure, visibility, and permeability. Whathappens to spatial genotypes when they aresubjected to such tactics, and does syntactic anal-ysis make sense when space is not clearlysegmented? It is our view that the analysis is worththe effort so long as it is coupled with a seriouswarning about the status of the diagrams. Thediagrams are not plans; they are designed to revealthe modes of access and control through the spatialstructure. The diagrams have a mimetic relationshipto the plan but are not mechanically derived from it.Boxes on the diagrams include both separate roomsand semi-separated spatial �elds. They are neces-

sarily interpretive, but they have an empirical basisin the �ows of movement through the buildings.For our purposes here, the diagrams should not beread mechanically but as �elds of sociospatialencounter.

The spatial structure is a “structuring structure”of what Bourdieu terms the habitus, the embodieddivisions and hierarchies between things, persons,and practices that construct the social world.15 Our“positions” within buildings lend us our “disposi-tions” in social life; the spatial “division” of ourworld becomes a “vision” of our world. The build-ings we inhabit, our habitat, our spatial habits, allreproduce our social world. Syntactic analysis opensup many questions. What kinds of agency areenabled and constrained by the particular buildinggenotype within which it is structured and whoseinterests are served? How is everyday life bracketedand punctuated into sociospatially framed situationsand locales? How does architecture frame the socialgaze through structured realms of visibility? Whatregimes of normalization are enforced and in whoseinterest?

Hillier distinguishes between “long” and“short” (or deep and shallow) models of interiorspace: the long model conserves and reproducesstatus and hierarchy, whereas the short modelgenerates new possibilities for social relations:

A ritual is a long model social event, since allthat happens is governed by rules, and a ritualtypically generates a precise system of spatialrelationships and movements through time. . . .A party is a short model event, since its objectis to generate new relationships by shuf�ingthem in space, and this means that rules mustbe minimized by using a spatial “short model.”. . . In a short model situation, space evolves tostructure, and often to maximise, encounterdensity.16

A permeable network of spaces and the openplan have long been linked to practices of social

freedom, yet any con�ation of physical enclosurewith social constraint, or of open space with liberty,is a dangerous one. Buildings are increasingly calledupon to produce an illusion of freedom coupledwith the reality of control and surveillance. Trans-parency may be used to couple a proliferating gazewith diminishing access.17 And freedom of associa-tion within a particular social group can build thesocial capital of that group vis-a-vis other groups.This is what Hillier and Hanson term the correspon-dence model of urban space in which spatial zones“correspond” to social groupings. High correspon-dence is when all those who share a spatial zonealso share a social label.18 A high level of correspon-dence is relatively deterministic of patterns of socialencounter: space operates to exclude randomencounter and to keep “difference” at a distance. Anon-correspondence system will mix people ofdifferent social identity; it is a spatial model thatbreeds encounter with difference. There is a sociallogic to the boundary between exterior and interiorspace in that interior space is more strongly struc-tured and segmented: the correspondence modelprevails with a primary function in the reproductionof social relations. This is countered by the non-correspondence model more common in exteriorspace. Each of these is in contradiction: randomencounter would undermine the social reproductivefunction of interior space, and the determinism ofinterior structures would kill urban diversity. This isprecisely what happens in urban space when tree-like genotypes (such as the housing enclave) invadepublic space. Koolhaas is attempting the opposite,playing with this tension between inside andoutside, using the encounter structures of urbanspace to effect innovations in interior space.

This attempt at a forced �t between spatial�elds and spatial syntax is surely a contentious issueand one which we cannot deal with at any lengthhere. The concept of space is very slippery andchanges meaning between disciplines and betweendifferent ideologies within architectural theory. Ingeneral terms, Hillier’s de�nition is more empirical

2. Educatorium, University of Utrecht

and materialist, whereas Koolhaas appears to adopta Deleuzian epistemology incorporating ideals of“smooth space.”19 Yet the work of both rests uponshared claims about the importance of socialencounter and shared concerns for architecture as“machine.” We now move to a spatial analysis oftwo of Koolhaas’s recently completed projects.

Factory of Learning:The Educatorium at UtrechtThe Educatorium for the University of Utrecht wascompleted in 1997; it houses a cafeteria, two largelecture theaters, and a cluster of examination

rooms. (See Figure 2.) According to the projectarchitect’s statement, it was conceived as the hub ofa campus servicing fourteen faculties and researchfacilities, a “rendezvous and exchange point,creating a new center of gravity,” which is to“embody the university ‘experience’: the socialencounters of the cafeteria space, the learningand exchange in the auditoria/classrooms, and theindividual rites of passage played out in the exami-nation halls.”20 There was a deliberate attempt togenerate diverse forms of social encounter inthe building: “seeking potential overlap betweenthe programs and encouraging exchange between

the users of its diverse functions, whilst allowing apragmatic and nearly autonomous use of individualspaces.”21 This architect’s statement introducesa series of key phrases and metaphors that havebecome the primary frameworks of critique in otherjournals.22 The characteristic blurring of inside andoutside in OMA projects is described through themetaphor of the “synthetic landscape.” The entry tothe building is described as a tilted ground planeand urban plaza that then continues as an interiorsloping “�eld” upon which the two auditoria areplaced like �gures in a landscape. This rising �oor-plate, which folds upwards and back to become the

Architecture and Freedom?Programmatic Innovation in the Work of Koolhaas/OMA

8

9 dovey and dickson

wall and then roof of the building, is described byKoolhaas as a “social magic carpet,” an urban land-scape of play and social encounter imported intothe architecture. (See Figure 3.) The �oor that foldsinto a wall has become the iconic image of thebuilding: one of the photographs provided forpublication shows a skateboarder “sur�ng” thecurved surface of the interior landscape. One of theauditoria has an entire wall open to the view and isdescribed as an amphitheater set in the landscape.Examination rooms are also described as interiorlandscapes that are able to be �exibly subdividedfor different functions. A permeable spatial struc-ture is deliberately designed “to act as a network inwhich students and users are free to discover theirown alternative shortcuts and to ‘drift’ through(the) building. Rather than attempting to dictateany particular pattern of use, the design of theeducatorium seeks to create a synthetic landscapeopen to individual choice.”23 Circulation areas aredesigned as a series of “pause spaces” forimpromptu hanging out between exams or lectures.

The spatial analysis diagram (Figure 4) shows abuilding that is accessible, shallow, and highlypermeable. The building is accessed publiclythrough eleven entry points from the exterior andother buildings. For a building of this size andcomplexity, this is a very shallow structure indeed:all major spaces are accessible within six levels ofdepth. With the exception of service spaces (whichhave been omitted for clarity), there are no dead-ends whatsoever. The building has three majorfunctional “attractors”: the auditoria, examinationrooms, and the cafeteria—each of which is coupledwith a major social circulation space. These threezones are organized vertically with the cafeteria onthe ground �oor, auditoria above, and the examrooms occupying the upper levels. The major circu-lation spaces and routes between them areunenclosed. (Spaces enclosed by doors are markedby dark frames on the diagram.) Four major foyersor “pause spaces” form a series from the rampedplaza on the exterior to the main foyer, which leadsupwards to the folded foyer and then back to a

balcony foyer outside the examination rooms.Although the plan has a high level of permeabilitywith a multiplicity of pathways, the main foyer alsooperates as a control space through which all of theopen circulation systems within the building pass.(See Figure 4.)

The Educatorium is repeatedly described in theliterature as a “factory of learning,” a phrase thatresonates with Koolhaas’s aesthetic, his machinemetaphors, and the role of the university as aknowledge factory. While knowledge is produced inthe research centers and staff of�ces deeper in theuniversity (in the spokes of the “hub”), fragmentsof this knowledge are revealed in the spectacle ofthe lecture theaters, discussed in the foyers andcafeteria spaces, and then examined in the enclosedrooms above. Markus has shown how the spatialstructure of the lecture theater surrounded by a�eld of highly permeable social space dates fromthe enlightenment; here knowledge is brought intothe light from a deeper source and discussed insocial space.24 Knowledge is legitimated in part bylocating its sources in deep spatial programs. Thecommunication from lecturer to student is reversed

in the examination rooms, in which studentsperform for lecturers and knowledge is tested underritual conditions of surveillance. This is what Hillierdescribes as the long model of spatial planningderived from the reversed spatial syntax of discipli-nary institutions of the enlightenment that placesubjects under surveillance deep within the spatialstructure. In the sense that this building is seen asthe “hub” of the university, one would expect to�nd staff of�ces and laboratories (the production ofknowledge) on the branches of the tree-like struc-ture. The spaces where student performance islegitimated are found deep within the hub. Here,the Educatorium becomes partially reversed,inducting its subjects into regimes of normalizationand surveillance in relatively deep space. The exami-nation zone is �ve to six levels deep within thebuilding; it does not receive the level of architec-tural attention of the shallow zones and does not�gure in the published photographs in magazines.Here the �eld of play stops and work begins.Although all examination rooms have multiplepoints of entry, they are each end points to spatialmovement. The shortest routes of access and egress

3. Educatorium, ground and �rst levels

(Photo by Christian Richters).

to these levels are not through the open foyersbut via the enclosed stair and the elevator. (SeeFigure 4.)

One can read the Educatorium as a radicallyinnovative building at its shallow levels with aconservative depth. The socialization of students,contact between students and staff, and thedelivery and sharing of ideas all take place in therelatively shallow network of social spaces. Yet thegrading of student performance, the legitimation ofinstitutional knowledge, remains deeply embeddedin the spatial structure. The two key metaphors of“synthetic landscape” and “factory of learning”re�ect the ways that the �eld relations of the land-scape have been imported into the factory tourbanize the building. Yet the synthetic landscapeof the folded �oor/wall/roof does not encompassthe examination rooms wherein the building moreclosely resembles the instrumentalism of the factory.Graa�and has suggested that Koolhaas’s work is asomewhat Faustian practice that embodies adialectic between the freedoms he seeks and thetree-like institutional structures in which such prac-tices are embedded.25 This entails a certain

acceptance of prevailing social and economic forces,the rejection of an architecture that simply resistsauthority in favor of a realpolitik wherein the desirefor the new is harnessed to make what one can in adif�cult world. The circulation system in thisbuilding is in many ways a masterful piece ofdesign, but it achieves this by integrating suchprogrammatic innovation with entrenched spatialgenotypes. Its freedoms of movement andencounter urbanize its interior, but only to the pointthat it does not threaten the knowledge/powerregime that produces the building in the �rst place.

Machine as Heart: The Houseat FloiracThe house at Floirac was completed in 1998 on ahilltop site outside Bordeaux in northern France andhas been widely published since then.26 The client isa family whose father was recently con�ned to awheelchair as the result of an accident, and thehouse was largely designed around his needs. Kool-haas suggests that it is not a house for an “invalid”but an architecture that denies “invalidity.” Thefamily’s previous house in the medieval section of

Bordeaux was described by Koolhaas as a “prison”and the father himself describes the new house ashis “liberation.”27

The house is organized with a total of fourvertical-movement systems connecting threeformally distinct �oors: a highly transparent living�oor sandwiched between the heavy mass of lower(kitchen/entry) and upper (sleeping) �oors. (SeeFigure 5.) The base level is an entry courtyard withcar access, framed by the house on one side withguest and servant quarters on the other. At thislevel, the house is excavated from the hillside andlikened (by Koolhaas) to a “sequence of caves” or“cellar” housing the entry, kitchen, wine cellar, andtelevision room.28 The middle level is a fully trans-parent, glass-enclosed slice of living/dining andgallery/study areas structured into one large �eld ofvisual and functional encounter. (See Figure 6.) Amotorized glass wall slides away to erase theboundary with the outdoor terrace, landscape, andcommanding views across Bordeaux. Just as theinterior is opened to the landscape, so the exterioris to be furnished with artworks using a specialtracking system in the ceiling. The bourgeoisdrawing room (once the “withdrawing” room) slidesout from the house. The bedroom accommodationon the top level is enclosed in a horizontal slab,pierced with porthole-sized windows, and designedto appear as if suspended above the transparentliving zone, like a sandwich about to collapse on itscontents.

The four vertical-movement systems are threestairways and an open elevator. The elevatorprovides the wheelchair access: a platform of 3 by3.5 meters that rises and descends on a hydrauliccolumn to align with each of the three �oors. Theplatform has no walls or balustrades, and itbecomes a part of each room it aligns with. AsKoolhaas puts it: “The movement of the elevatorchanged, each time, the architecture of the house.A machine was its heart.”29 At the ground �oor, theplatform becomes an alcove off the entrance/kitchen and provides the access to a wine cellar.At the middle �oor, it becomes an unenclosed part

4. Educatorium spatial analysis.

Architecture and Freedom?Programmatic Innovation in the Work of Koolhaas/OMA

10

11 dovey and dickson

5. House at Floirac, Bordeaux plans.

6. House at Floirac, middle �oor (Photo by Christian Richters).

of the living/dining areas with views out to thelandscape and terrace. At the top �oor, it becomesan alcove off the father’s bedroom. At each level,the platform slides across bookshelves that lineone side.

The four vertical-movement systems generatea highly interconnected spatial structure for thelower �oors. (See Figure 7.) However, they are alsoorganized for specialized use: the mother’s stair tothe east, the guest stair near the entry, and thechildren’s stair to the west. Koolhaas regards theelevator as the key liberating technology of our era,and here it is the elevator platform that renders thehouse accessible by wheelchair. However, unlike theurban elevator, which is shared by wheels and feet,here it is the domain of one person because use byothers would leave the father stranded. Thismachine is the “heart” of the house, where it trans-forms the architecture and places the father incharge. The father controls the architecture, and theposition of the elevator becomes a signi�er of hispresence and absence. When the father is out, onthe ground �oor, or in bed, then the main livingspace is left with a central void. This void has no

handrails and at least one reviewer has suggestedthat the space produces a sense of genuine insecu-rity and risk.30 This void (reminiscent of Eisenman’sfamous hole in the �oor between the twin beds)can be read as a deconstructive challenge to thetraditional idea of house and home as reinforcingontological stability. Yet the void is created anderased in a gender-speci�c manner: the centralliving space of the house is secure only when thefather is present.

The elevator is furnished in some photographsas a study with a desk and lamp. Because it is linedwith bookshelves and controls all access to the winecellar, it can be interpreted as a reconstruction ofthe male “den” brought into the light and mobi-lized, transposed from deep to shallow space withthe walls removed. However, when the platformretreats to the top �oor, it seals the gap to create afully enclosed space, deep within the spatial struc-ture. This adult bedroom zone is structured in along loop with the two bedrooms at once separatedand connected by a bathroom and a balcony. (SeeFigure 7.) The children’s bedrooms form a moretraditional tree-like syntax on the same level but

entirely severed from the adult zone and inacces-sible to the father. Surveillance over children is theonly function not afforded the father.

The house at Floirac is in many ways a recon-struction of the bourgeois house with its servantquarters and cellar dug into the hillside, surmountedby the piano nobile with its commanding views, andthe attic story with its tiny windows. The house canbe construed as a play on the Bachelardian arche-type, itself �rmly based in the French bourgeoishouse, with its cellar and garret framing theeveryday life of the middle �oor.31 (See Figure 7.)However, it is a radically innovative and imaginativehouse, both formally and spatially. It combines arethinking of the dialectics of inside/outside (as inMies’ Farnsworth house) and vertical/horizontal(Corbusier’s Villa Savoye), but with greaterprogrammatic dynamism and complexity (as in theRietveld Schroder house). Although the structure ofthe house is highly ringy on the lower levels, it isalso conceptually tree-like with the elevator as itsstem. It embodies new forms of both liberation andsocial control, and gender divisions are enhancedrather than challenged. As Koolhaas puts it, “amachine is its heart,” and the machine is a patriar-chal prosthesis. Although positions could betransposed (with a woman controlling the space),the structure of the house would remain hierar-chical. Is this new spatial hierarchy an accidentalbyproduct of Koolhaas’s obsession with theelevator? Or is it a deliberate tactic of bringingauthority into the light rather than resisting it,exposing the “nerve,” as Kipnis puts it. In eithercase, it seems a dangerous move.

CommentLiberty is a practice . . . [I]t can never beinherent in the structure of things to guaranteethe exercise of freedom. The guarantee offreedom is freedom.

Foucault32

So what can be made of Koolhaas’s desire tounhinge architecture from its role in social reproduc-

7. House at Floirac spatial analysis.

Architecture and Freedom?Programmatic Innovation in the Work of Koolhaas/OMA

12

13 dovey and dickson

tion? Koolhaas’s goals are generally laudable: hetreats interior space as a �eld of play that resistsany simple mimetic relationship with social struc-ture, and a permeable spatial network is a primarydesign tactic. He wants to defy the “social logic ofspace,” to free up the programmatic imperativesthat lock architecture into the service of a highlychoreographed and ritualistic reproduction of sociallife. Yet the freedoms of which he speaks remainsassertions that cannot be easily tested. Kipnissuggests that, “For Koolhaas, architecture is able. . . to engender provisional freedoms in a de�nitesituation, freedoms as the experiences, as thesensations, as the effects pleasurable, threatening,and otherwise of undermining select patterns ofregulation and authority.”33 What this suggests isthat such claims need to be considered in thecontext of everyday experience and social practiceenmeshed in the micropractices of power and liber-ation that infuse everyday life. Yet such critique ofplace experiences is one that Koolhaas explicitlyeschews.34

Koolhaas’s work is brilliantly innovative, but itis not always what it seems. In her critique of theFloirac house, Colomina suggests that he operatesin the mode of a magician, distracting the eye withone hand, concealing what he is up to with theother.35 Koolhaas’s formal inventiveness distractscritical attention from his programmatic surgery,which at times constructs illusions of “freedom” thatcan conceal what has not changed. Programmaticinnovation can be reduced to signi�cations of prac-tice. Koolhaas has indeed alluded to himself as amagician producing “sublime moments of illusion,”and there is no suggestion here that the “magic”does not work in certain ways.36 What is missing,however, is an understanding of freedom as a formof practice: something people do rather thanconsume. Koolhaas does indeed challenge theprimary genotypes of sociospatial reproduction, yetat the same time he generates illusions of an archi-tecture that has been freed from spatial ideology.And these illusions can be a cover for new practicesof power or for more of the same.

Notes

1. Jeffrey Kipnis, “Recent Koolhaas,” El Croquis 79, (1998): 26–31,

p. 27.

2. Alejandro Zaera and Rem Koolhaas, “Finding Freedoms: Conversa-

tions with Rem Koolhaas,” El Croquis 53 (1992): 6–31.

3. Bill Hillier, Space Is the Machine: A Con�gurational Theory of Archi-

tecture (Cambridge, UK: Cambridge University Press, 1996); Bill Hillier

and Julienne Hanson, The Social Logic of Space (New York: Cambridge

University Press, 1984).

4. Arie Graa�and, “Of Rhizomes, Trees and the IJ-Oevers, Amsterdam,”

Assemblage 38 (1998): 28–41, p. 36; Bart Lootsma, Superdutch: New

Architecture in the Netherlands (London: Thames and Hudson, 2000),

p. 181. This account of Koolhaas’s work also relies strongly on Zaera

and Koolhaas, “Finding Freedoms,” and Kipnis, “Recent Koolhaas.”

5. Kipnis, “Recent Koolhaas,” pp. 29–30.

6. Ibid., p. 30.

7. Fredric Jameson and Michael Speaks, “Envelopes and Enclaves: The

Space of Post-Civil Society,” Assemblage 17 (1992): 31–37, p. 33. See

also Graa�and, “Of Rhizomes.”

8. Stan Allen, “From Object to Field,” Architectural Design 127 (1997):

24–31.

9. Ibid., p. 24.

10. Ibid., p. 31.

11. Hillier and Hanson, The Social Logic of Space; Hillier, Space Is the

Machine.

12. For a more detailed theoretical account of this adaptation of spatial

analysis, see Kim Dovey, Framing Places: Mediating Power in Built Form

(London: Routledge 1999), ch. 2. For other uses of spatial analysis, see

Thomas Markus, Buildings and Power (London: Routledge, 1993); Julia

Robinson, “Messages from Space,” in Roberta Feldman, et al. (eds.),

Power by Design (Oklahoma: EDRA Proceedings, 1994), pp. 145–151.

13. Dovey, Framing Places, pp. 22–23.

14. See Hillier and Hanson, The Social Logic of Space ; Markus, Build-

ings and Power ; and Dovey, Framing Places, ch. 2.

15. Pierre Bourdieu, Outline of a Theory of Practice (London: Cambridge

University Press, 1977), ch. 1; Pierre Bourdieu, Pascalian Meditations

(Cambridge, UK: Polity Press, 2000), ch. 4.

16. Hillier, Space Is the Machine, p. 7.

17. Markus, Buildings and Power, p. 18.

18. Hillier and Hanson, The Social Logic of Space, pp. 255–261.

19. G. Deleuze and F. Guattari, A Thousand Plateaus (London: Athlone,

1987), ch. 14.

20. Christian Cornubert, “Educatorium, Utrecht, Holland,” Domus 800

(Jan. 1998): 42–47, p. 43.

21. Cornubert, “Educatorium,” p. 44.

22. Peter Buchanan, “Rem Koolhaas/OMA: Educatorium at University

of Utrecht,” Architecture and Urbanism 336 (1998): 24–45; F. Irace,

“Educatorium: OMA a Utrecht,” Abitare 379 (Dec. 1998): 104–114;

Raymund Ryan, “The New Dutch School,” Architecture 87/3 (1998):

136–143; and Connie Van Cleef, “Campus Landscape,” Architectural

Review 1225 (March 1999): 50–53.

23. Cornubert, “Educatorium,” p. 45.

24. Markus, Buildings and Power, p. 230–233.

25. Graa�and, “Of Rhizomes.” See also Kim Dovey, “Multiplicities and

Complicities: Signifying the Future at Euralille,” Urban Design Interna-

tional 3/3 (1999): 89–99.

26. OMA Rem Koolhaas LIVING (Bordeaux: Birkhauser, 1998); Beatriz

Colomina and Blanca Lleo, “A Machine Was Its Heart: House in Floirac,”

Assemblage 37 (1998): 36–45; Colin Davies, “Machine for Living,”

Architecture (Dec. 1998): 72–82; M. Emery, “House at Floirac, Gironde,”

Architecture d’ Aujourd Hui 320 (Jan. 1999): 24–33; and M. Nesbit,

“House Play,” Artforum 37/3 (1998): 94.

27. Colomina and Lleo, “A Machine Was Its Heart.”

28. Emery, “House at Floirac.”

29. Quoted in Colomina and Lleo, “A Machine Was Its Heart,” p. 42.

30. Davies, “Machine for Living.”

31. Gaston Bachelard , The Poetics of Space (Boston: Beacon, 1969).

32. Michel Foucault, “Space, Power and Knowledge,” in Neil Leach

(ed.), Rethinking Architecture (London: Routledge 1997): 367–379,

p. 371.

33. Kipnis, “Recent Koolhaas,” p. 27.

34. Rem Koolhaas and Bruce Mau, Small, Medium, Large, Extra-Large

(New York, Monacelli Press, 1995); Rem Koolhaas, “No Grounds Against

a Non-Place,” in Euralille (Basel: Birkhauser, 1996): 51–71.

35. Colomina and Lleo, “A Machine Was Its Heart.” The magic seemed

to work on this critique which is curiously gender blind.

36. Koolhaas, “No Grounds Against a Non-Place,” p. 190.