The diagram as abstract machine

Abstract

Whether graph or chart, the architectural diagram is today an ubiquitous presence. As graphic inscription of abstraction in space, since the 1990s, the notion of diagram has been so much extended that now it nearly encompasses every aspect of design. To think of the diagram as an architecture of ideas (or, more classically, the idea of architecture) means to be still ensconced in some sort of platonic conceptions (Garcia, 2010). To avoid this trap, a first step would be to turn to Gilles Deleuze’s notion of the diagram as an abstract map, and to show how the model acquires its meaning, specifically when confronted with biological paradigms. Such understanding may lead to a better comprehension of the present algorithmic nature of diagrams. These scripted procedures refer to form, or, more precisely, to processes of morphogenesis. Their aim is to enhance a modulation between natural components, physical elements, and architectural design. A range of practices (or protocols) based on adaptable (customable) software, capable of producing changing modalities of a structural topology driven by performance, are currently available (Teyssot and Bernier-Lavigne, 2011). For instance, addressing the issue of the use of genetic algorithm in design, Manuel De Landa, inspired by Deleuze’s work, has proposed to introduce three theoretical levels of complexity: to think in terms of population (not the individual); to think in terms of differences of intensity (thermodynamic and entropic); lastly, to think in terms of topology (De Landa, 2002). The question addressed here will be therefore to know if (and how) the diagram is able to topologise the various fields of design.