Back to basics
Richard Weston is one of the liveliest writers on architecture. He combines acute readings of buildings that he likes with an infectious enthusiasm for the subject as a whole. He has his favourites, which generally means architects who treat their trade as an exercise in building technique, like Aalto or Utzon, rather than a vehicle for theoretical speculation. It means that Weston has a blind spot for wider cultural interpretations of architecture, and he also doesn't care for new digital media. He much prefers just looking at the evidence of actual buildings, or drawings of buildings.
So the intention of his new book, which gives us plans, sections and elevations of 106 key buildings from the 20th century, is as a kind of Modernist version of Banister Fletcher. Each entry gets a double-page spread, including a brief written commentary and one or two thumbnail photographs. There are no perspectives or other forms of threedimensional projection. Acknowledged 'masters' such as Le Corbusier, Wright and Mies get plenty of look-ins. When it comes to design approaches that Weston has less sympathy for, such as Post-Modernism or Neo-Classicism, they have a token presence, mainly through curiosities such as Josef Plecnik's Church of the Sacred Heart in Prague.
The question is whether Weston's 'back to basics' approach works. His method appears to start from the assertion of Alberti and others, echoed by Le Corbusier, that three-dimensional representations cannot quite be trusted, since they lack the clear measurable precision of two-dimensional drawings. Weston doesn't mention the motive behind his book, but one suspects that it was frustration about the way that architectural students now find information about precedents through the internet. This offers lots of easily accessible photographs and three-dimensional digital renders, but no real or systematic idea of how a building is put together, or how a user actually moves around it in sequence.
Yet there is an inherent contradiction in Weston's solution. Given that the information is not standardised, and because almost no one draws by hand any more, all of these 106 buildings have been redrawn from scratch in plan, section and elevation by a sizeable team using CAD software. The benefit is that a CD of the drawings can be supplied with the book; the perversity, of course, is that it is precisely in designing and visualising three-dimensional spaces that computers are so superb.
It is in this conscious emasculation of the computer that problems arise. Weston's approach works best for canonical Modernist buildings whose configuration is already well embedded in our minds, like the Villa Savoye. In other words, it is fine for buildings of reasonable spatial complexity that have defined floor levels and spatial volumes - even when, as in Wright's Robie House, these might partially dissolve into each other.
But in the case of more recent buildings, difficulties abound. First are designs in which the plans, sections or elevations are so simple as to be mute if presented in themselves; I would defy anyone to appreciate the nuances of Herzog & de Meuron's Goetz Collection in Munich, as represented so crudely here.
Then there are instances where the difficulty stems from the decision to compress the plans and sections of a genuinely complex building into the tight two-page format. Fans of the Kunsthal in Rotterdam will know that Koolhaas/OMA needed at least six floor plans (or, more accurately, horizontal sections) to map out the interweaving circulation ramps. In this book there is just one tiny section and three plans on which too much information has been superimposed or, for certain elements, strangely omitted. If they are read literally, the drawings here simply do not explain the building, and one has to rely on substantial foreknowledge.
More problematic still are cases where those involved in making the drawings do not seem to know the building, and have been inconsistent or have made mistakes.
The most misleading entry is Siza's Museum of Contemporary Art in Santiago de Compostela, a building of such subtlety that you cannot really grasp it until you study it in the flesh. The plans here are a mess: staircases are shown rising the wrong way; solid elements are shown as gaps, and vice versa; rooms do not have doors; some circles that read like columns are no such thing; the basement gallery suite is not shown at all; the roof terrace levels are not indicated correctly; and worst of all, the single most brilliant spatial move has been omitted from the key plan.
This is where the first-floor galleries, with their famous upside-down-table light devices, are suddenly cut away at one point to be revealed to the ground floor gallery in a double-height space. Siza extends a thin bridge across this volume to give access to a pop-out glazed door (typically, he refrained from providing any handrails, meaning that the bridge was closed off when I went there, as it is lethal). Is this properly explained in this book?
Perhaps this is being too picky. But it does seem to disprove Weston's assertion that only through plans, sections and elevations can we truly understand architectural volumes. If a three-dimensional digital model had been produced of Siza's design, many of the errors here would have been spotted. For all of his fervent proselytising, and the sincerity of his aims, Weston is fighting a losing battle. Yes, there is still great value in two-dimensional drawings, but it is computer visualisations that are making the headway now.
Murray Fraser is a professor at the University of Westminster