Alvaro Arancibia Tagle, Cité Housing in Santiago de Chile (AA Projective Cities, 2013)

Introduction

DESCRIPTION, ANALYSIS AND TRANSFORMATION

DRAWING AS DIAGRAM

Drawing is the basic means to communicate and reason an architectural or urban design. During the design process, drawings are produced that 'describe', 'analyse' and 'generate' ideas and structures. Yet in architectural drawings, these three modes overlap and are often difficult to distinguish. Description of architectural designs through graphic representation always requires some abstraction of reality and is therefore partially analytical. Analysis itself depends on conventions of description and often a comparison to established norms. And a generative drawing frequently is the conclusion of a comparative analysis, especially when deriving from a design method based on case studies. A sketch then can be descriptive, analytical and generative, depending on its context. In fact, context is critical to all drawings. We read a drawing differently when it is presented by students to explain a speculative project, is used by designers and clients to discuss how quantitative or aesthetic requirements are met, or is part of a tender package detailing building parts and elements for construction. Different drawing conventions have thus developed that are suitable to each context.

Defined by the interactions between description, analysis and generation, architectural and urban drawings are diagrammatic reductions that generally convey abstract spatial relationships and orders - which can serve as blueprints for building. To materialise physically what are, despite representing a 'reality', essentially conceptual ideas, translation is required, as a diagram substantially differs from the reality it describes. But whether a project is meant for realisation or not, the principle of conceptual and material transformation is part of all design relying on representation and its translation.1 Any spatial concept is generative and can always be realised at different scales and in different materials, in each case demanding a new analysis and set of drawings, details, construction methods, etc. Therefore, the generative can be defined in these terms as a transformation that follows from analysis. It is not simply an artistic vision. Drawing Architecture and the Urban is structured around the three principal concepts of drawing as a means of description, analysis and transformation.

The need for abstraction in architectural drawings is intensified by computer-aided design (CAD), as the idea of a single drawing is obsolete, and potentially an infinite number of drawings are produced from the same file. What seems a simple observation of a mechanical process has a significant effect on the way we can instrumentalise drawing and conceive its role. A drawing produced by CAD requires continuous rethinking of scale, detail and graphic design. While hand drawing emerged within a tradition of imitative representation, computer drawing is associated with a generative process of abstraction, in which generalisation and its transformations have become operative for design and its description - a fact that is recognised in the building industry through the increasing use of building information modelling (BIM). Thus with CAD, the conventions and problems of drawing have changed from the logic of hand drawing. The inherently generative nature of CAD requires a new judgement as to what and how we draw - indeed, even as to how we conceive a project. The great potential of the computer drawing and model partially explains why, during its rise in the 1990s, an entirely three-dimensional and self-referential formal generation was pursued that deliberately avoided being analytical. Computer generation was turned into an end. Yet this book is a return to an analytical approach. It argues through the drawings themselves that graphical thinking is a central means of architectural reasoning, one which with the advent of the computer has established not only new forms of analysis and graphic knowledge but also a new aesthetic and style.

The three functions of diagrams to describe, analyse and transform refer respectively to three more historical architectural concepts: convention, interpretation and invention. By following conventions - agreed meanings of representation and order established by practice - it is possible to describe and communicate clearly. Drawing is in this case a learned skill obtained through practice. Analysis on the other hand is interpretative, but relies on conventions of description and at times becomes the basis of invention through conceptual and formal transformation. The skill of analysis is part of the training of an architect. Yet the relationships between the six concepts directly reflect on changing practices and disciplinary knowledge, and are therefore continuously changing themselves. This is apparent when looking at the larger historical and theoretical context in which problems of drawing shaped the architectural profession and its approach to design. It transformed a descriptive drawing into a generative diagram and developed a comparative study of precedents into a design method while introducing the idea of a typological transformation. These concepts of the diagram, comparative study of precedent and typological transformation, are explored in the different chapters of this book.

FROM DESCRIPTIVE DRAWING TO DIAGRAMMATIC DESIGN METHOD

In the second half of the 17th century, the problem of accurate description through drawings was a major motivation for early archaeological studies of classical Roman precedents, with the aim to clarify the proportions of architectural orders.2 These surveys uncovered the incommensurability of proportions in different classical buildings, which eventually led Claude Perrault to conclude in his Ordonnance for the Five Kinds of Columns after the Method of the Ancients of 1683, that proportions are relative. He resolved the problem of comparing different proportional dimensions by introducing a new convention of measurement that permitted him to express proportions as whole numbers and not fractions (figure 1).3 Following Perrault's empirical approach, notably Pierre Nativelle in his Nouveau traité d'architecture (New Treaty on Architecture, 1729) proposed a new graphic abstraction. He radically eliminated the differences between architectural orders by conceiving them as an abstract and generic scheme for design (figure 2). This transformed a previously descriptive system of drawing into a diagrammatic matrix of possibilities.

While Nativelle introduced a new diagrammatic thinking, at the same time he diminished the role of precedents, which Jacques-François Blondel in contrast saw as the foundation of architectural practice and training. Blondel introduced the case-study method from law, and understood architectural precedents only as meaningful when maintaining relevance to practice by providing principles on which similar current design problems could be judged and decided on.4 Precedents were thus defined as potentially a historical. His method required comparative drawings to analyse case studies, and redefined invention as a modification and recombination of received forms. Blondel's position further necessitated a clarification of the role of history in the process of design and its analysis. This was provided by Julien-David Le Roy in The Ruins of the Most Beautiful Monuments of Greece (1758), who was the first to formally introduce history as a counterpart to theory in architecture. Le Roy posited that without accepting historicism, architecture could not be freed from the constraints of first origins - by which he meant classical Greek or Roman styles. By introducing the concept of history, it could be separated from the practice of architecture, which was then considered its theory. Complementary to his writing, Le Roy instrumentalised drawing, especially in two pivotal diagrams. The 'Plans of the most remarkable churches built from 326 to 1764' showed Jacques-Germain Soufflot's design for Sainte-Geneviève as the synthesis of three formal developments: the cross-shaped plan, parallel rows of free-standing columns in the basilica and the dome (figure 3). Although the diagram appears simply comparative, it suggests that a typological progression and transformation underlies the judgement of design, permitting one to recognise Soufflot's plan as superior. The second important diagram similarly compared a progressive transformation, that of Egyptian and Phoenician, Greek, Roman and Christian huts into temples with increasing scale and complexity (figure 4). The graphic comparison proposes a clear typological sequence of transformation, and is an explicit example of a typological and diagrammatic reasoning, as it relates generic and specific form in order to construct a lineage and classification of a group of buildings.

Figure 1: 'Plate 1' showing the comparative proportions of the five architectural orders (Tuscan, Doric, Ionic, Corinthian and Composite) as measured by the petit module, in Claude Perrault, Ordonnance des cinq espèces de colonnes selon la méthode des anciens (Paris: Coignard, 1683)

Figure 2: 'Les cinq Ordres d'Architecture dans une mème hauteur suivant Vignole' [The five Orders of...