explores how lists operate as diagrams, as persistent tropes in literature, from Jorge Luis Borges’ fictional Chinese encyclopedia (1942) to Georges Perec’s lists and inventories (1974). Lists are adhoc collections or inventories in random sequence, that seek to be exhaustive while remaining open-ended; alphabetical lists such as dictionaries create new meaning through unexpected adjacencies and juxtapositions. SMLXL, authored by Koolhaas, Mau and OMA (1995), uses competing ordering systems and lists, alphabetical, scalar and temporal lists to weave together a methodological field of citations, references, literary genres and cultural languages. The chapter examines diagrammatic operations underpining the multi-vocal narratives and multi-focal layouts of SMLXL’s book world and the homologous spaces of Koolhaas’ architecture and urbanism.
Chapter five, on Taxonomies and Typologies, examines those contrasting systems of classification as diagrams used to generate architecture, beginning with the opposing positions of J.N.L. Durand and Gottfried Semper. Durand’s methodology (1805) plays out on an abstract grid inviting the designer to freely combine architectural types and typological elements that thus are detached from their historical and cultural context. Semper’s counterproposal sought a method of design modelled on biological evolution (1853, 261). His taxonomical tree situates artefacts in familial relationships to each other, analogous to the evolution of species. Both of these competing generative diagrams, undergoing a series of reinterpretations have gained new relevance in the context of the early 21st century, manifest in the architectural practices of Forreign Office Architects (typology) and of WORKac (taxonomy). Drawing in the Deleuzian notion of diagrams as abstract machines, Manuel de Landa’s concept of a genetic algorithm in architecture builds on Semper’s taxonomy, and questions architects’ agency within design methodologies and scenarios of allopoïesis and autopoïesis.
The final chapter, on Rota and Network, extends this exploration of poïesis into the realm of Utopian thought, to cosmopoïesis. While not annotated with diagrams, Thomas More’s description of Utopia (1516) paraphrases contemporaneous worldviews embodied in rota diagrams. Notions of cyclical time, of hierarchical stratification between core and periphery are symbolised in a series of concentric circles evoking stability as well as rotational movement. The chapter contrasts Utopia against Agronica (1994), Andrea Branzi’s project for a weak urbanism constituted by a pervasive network, explained through a three-dimensional model that simulates infinite space in a mirror-box. In each case, Utopian proposals articulate their authors’ critique of an existing social, political and spatial system, while their use of culturally meaningful diagrammatic conventions, vitally interlinked with cosmography, reflects a prevailing world-view of their historical era.
I would like to thank Sigrid Loch, Alexandru Mălăescu, lulia Frățilă, Lara Rettondini, Oscar Brito, Sophia Psarra, Almudena Cano, Íñigo Cornago, Ed Wall and Tim Waterman for many debates and for their generous advice over many years, Anthony Vidler for his insightful critique of my PhD thesis. Special thanks go to Rochus Hinkel for inviting me to publish this monograph and for his guidance and help in crafting it.
Poché and Free Section
The test of a first-rate intelligence is the ability to hold two opposed ideas in mind at the same time and still retain the ability to function (Fitzgerald 1936, 41).
Within a vast arsenal of architectural techniques employed by OMA in their first decade, poché occupies a unique position, for two reasons. First, unlike design techniques adopted from Surrealism, such as the paranoid-critical method or the cadavre exquis, or metaphors such as the medical term lobotomy, the concept of poché is drawn from the history of architecture. Second, while appropriation to architecture of techniques originating elsewhere figures prominently in Rem Koolhaas’ theoretical output, the use of poché is never mentioned by Koolhaas or Zenghelis during OMA’s first decade. Only in 1999 Koolhaas finally acknowledged “a fascinating condition to work for the first time with so-called poché” (Oswalt and Hollwich 1998, 12-22), on House Y2K and the Casa da Música in Porto, thereby denying the apparent role of poché in the strategy of the void for the new town of Melun-Senart (1987) and in the project for the Très Grande Bibliothèque (1989). OMA’s ambiguous reception of poché during its first decade can be summarized as negation in writing alongside appropriation in design. It is as such a reaction to Robert Venturi’s extrapolation of the Beaux-Arts conception of poché to urbanism; Koolhaas has described Venturi as both inspiration and threat (2004, 150). Koolhaas has acknowledged: “I think that for instance the historicists very legitimately have accused modernists of being stupid about many things. And I think that in that sense, on an almost pragmatic level, I would say, yes of course there is a lesson, because now it is possible to be a better modern architect, simply because of their critique. You can incorporate your critique in your own things” (Koolhaas 1983). Repudiation of Venturi’s and Colin Rowe’s contextualist definition of poché acts as a polemic protective shield which allowed OMA to amalgamate poché with its tectonic antithesis, the free section.
Poché
In Beaux-Arts education, poché denoted the hatching or rendering in fields of colour of masonry that is sectioned in plan, which was applied to presentation drawings, but not to working drawings. Nevertheless, poché is as much a tectonic as it is a drawing convention, denoting load-bearing masonry construction which presumes space and structure to be congruent, in opposition to the free plan theorized by Le Corbusier in 1926. That same year, the Beaux-Arts theorist John F. Harbeson emphasized that “poché always encloses rooms” (1926, 188), which applied not only to the primary spaces bounded by walls, but also spaces contained within the hollow walls.
The theme of the “hollow wall” is a longstanding trope in the work of Koolhaas, beginning with the cells inserted into the walls of his 1972 thesis design at the Architectural Association, Exodus (Koolhaas 1977, 328-29), continuing with the 1974 House in Miami (Koolhaas and Spear 1977, 352), where “service areas such as pantry, powder room, bar and bathrooms are located within the thickness of the wall,” and the Story of the Pool (Koolhaas 1977, 356), its basin bordered by two thick, hollow walls accommodating locker rooms. The dominant impulse of these early Koolhaasian walls is to divide rather than enclose space, betraying their derivation from Koolhaas’ 1971 study The Berlin Wall as Architecture (Koolhaas 1995, 236). The early Koolhaasian walls act as radical disjunction, in opposition to Modernist orthodoxy postulating that “the inside should be expressed on the outside” (Venturi 1966, 70).
Urbanism and Contextualism
Robert Venturi, in Complexity and Contradiction, noted that “contradiction between the inside and the outside may manifest in an unattached lining which produces an additional space between the lining and the exterior wall,” and that “the space left over by this contradiction was taken care of with poché” (1966, 70). In 1968 and 1972, Venturi and Denise Scott- Brown extrapolated the spatial conception of poché to the scale of the city; they observed that “Nolli’s map of the mid-18th century (Figure 2) reveals the sensitive and complex connections between public and private space in Rome” (1968, 128). Their ideas were received with particular interest at Cornell University, by both Colin Rowe and by O.M. Ungers, with whom Koolhaas had in 1972 taken up studies. Following Venturi’s line of thought, Rowe, in Collage City, defined poché at two scales. First, at urban scale “a building itself may become a type of poché, (…) a solid assisting the legibility of adjacent spaces,” able “to engage or be engaged by adjacent voids, to act as both figure and ground” (1978, 79), and second, at the scale of building and façade, “ideal types” are adapted to and modified by “empirical context” (1978, 106) with poché acting as a technique of mediation.
Koolhaas sharply distanced himself from Rowe’s historicist tendencies and approaches, polemically deriding his “contextualist epiphany,” and criticizing that “the modern contextualist is forced to telescope vicissitudes of centuries into a single moment of conception” (1980, 48). His position drew on
