Arguments for tradition in architecture: 3

I wrote earlier about David Watkin’s criticism of Pevsner in ‘Morality & Architecture’. Watkin says much less about another advocate of modernism: Sigfried Giedion. But with Giedion, Watkin’s general thesis—that modernists have all signed up to an unreflective Hegelianism and a ‘belief in progress’—comes closer to the mark. In ‘Space, Time & Architecture: The Growth of a New Tradition’ (Harvard, 1941) Giedion does describe a Zeitgeist (and calls it that) and declares his obligation to bring it to light, for general benefit. Still, there is more going on with ‘Space, Time & Architecture’ than you get from Watkin. ‘Morality & Architecture’ promotes the idea of style, where this term is understood as naming collections of building features: patent features that are primarily understood visually, and which tend to be seen in characteristic families or groups. Giedion rejects the term style (and Watkin reacts to this rejection). But why does Giedion reject style?

‘Space, Time & Architecture’ offers, in place of styles, a distinction between what Giedion calls “constituent facts” and “transitory facts”. By ‘facts’, Giedion means the (factual) occurrence of building features; a “constituent fact” is the persistent appearance, over time periods that may be quite extended, of a certain feature. For example, an “undulating wall” is, for Giedion, a constituent fact. Watkin mistakenly thinks Giedion intends by “undulating wall” a modernist stylistic trope and assigns to Giedion a simple preference for it. But Giedion traces his undulating wall from the Italian Baroque (perhaps Borromini’s S. Carlo alle Quattro Fontane, 1638-41), through “the great dwelling complexes of the eighteenth and nineteenth centuries” into the present day (of the 1930s-40s); the undulating wall must then be understood as a feature that, if not ‘transcendent’ of style, is at least one that cannot easily be captured within a style definition. A better understanding of Giedion, I think, is that his “constituent fact” is an expression of knowledge how: that is, there is a known way to build curved walls; a geometric understanding, a capacity to make drawings on the basis of that understanding, and finally a capacity to transfer that information for use at full scale by stonemasons (or other fabricators). At one time that knowledge how did not exist, and now it does. This is an achievement; a newly embedded element within our institutions and accepted practices. It is this status of achievement that gives the “constituent fact” its rootedness, or persistence. A “transitory fact”, by contrast, involves no such achievement. The features in play with a “transitory fact” are mundane and no special—or new—knowledge is involved in their application. Giedion’s example is “furniture of the Second Empire in France”, typified by an eclectic re-mixing of classical motifs. The implication is that Second Empire furniture is something that did not happen earlier, but could have happened earlier. At the time that it happens, it is merely a novelty. As such, it may just as easily disappear again.

Giedion’s focus on persistence in ‘Space, Time & Architecture’ is reflected in the subtitle he gives the book: ‘The Growth of a New Tradition’. Some things become established in architecture. Let’s look at another example. Take the conoidal structure that’s hidden inside the dome of Wren’s St Paul’s Cathedral. This involved both a geometric and structural innovation. However, Wren himself had a conservative aesthetic. In Parentalia he writes:

“There are only two beautiful Positions of strait Lines, perpendicular and horizontal: this is from Nature, and consequently Necessity, no other than upright being firm. Oblique Positions are Discord to the Eye, unless answered in Pairs, as in the Sides of an equicrural Triangle: therefore Gothick Buttresses are all ill-favoured, and were avoided by the Ancients, and no Roofs almost but spherick raised to be visible.”

And:

“The misapprehending World measures the Excellence of things by their Rarity, or Difficulty of Framing, not by the Concinnity and apt Disposal of Parts to attain their End by a right Line as it were & the Simplest way.”

This, I’d suggest, explains the concealment of the part of the dome of St. Paul’s—the actual masonry shell—that carries the main forces from the cupola: for Wren, it was the right shape (roughly, a revolved catenary) to do the job of carrying imposed loads, but the wrong shape to be seen. It was therefore hidden by a roof structure of a more spherical appearance, executed mostly in timber.

Today, would a conoid still be the wrong shape to be seen? No; but what has made the difference? Giedion says that there is an additional dimension to innovation beyond architectural practice; that is, experiment in visual arts. For Giedion, experiment in fine arts—implicitly, quick experiment, at low cost—allows the application and open exhibition of new construction techniques. There is a gradual evolution of general taste, through new sculpture and painting, and in these changed conditions the new way of building can emerge. To illustrate this further claim, Giedion considers the output of the Swiss engineer and architect Robert Maillart (1872-1940); the twenty-six pages devoted to Maillart constitute one of the most rewarding parts of ‘Space, Time & Architecture’. Maillart’s main contribution, according to Giedion, is to shift the domain of structural analysis from planar elements—beam, column and arch—to surfaces; shells and vaults. This innovation is accompanied by the take up of reinforced concrete as a construction technique, with particular exploration of connection details: just how the steel reinforcement—the component that deals with tension forces—transitions from slab or vault to column is crucial to the realisation of greater plasticity. (Giedion is careful to illustrate these details. They are not incidental: they are part of the knowledge how.) Now vaults and shells—structural surfaces—do predate Maillart: Wren’s conoid is an example. Some of these early (typically masonry) structures even feature tension elements. However, along with his presentation of Maillart, Giedion cites the work of Impressionist and Cubist artists: this body of work—he argues—emphasises surface. In the context of an improved public appreciation of surface, a designer such as Maillart can now propose unornamented, pure forms. The shape of the optimised structural element is now aesthetically sufficient; nothing now needs to be hidden.

Of course, we can dispute this, or some part of it. We can look, for instance, at the French tradition of surface expression in stonemasonry (stereotomy). But still, what Giedion sets out here looks to be some sort of sociological model, or perhaps a kind of anthropology, and his case for acceptance of it is convincing. This is how people act, he says, and here is an example that has happened. Take this example and all of the related activity together, and we have something we have to call a new tradition. It exists. Here, Giedion is in descriptive, rather than normative mode: if he is wrong, it is not because we think we don’t have to do likewise in our own design practice; he is wrong because what he describes simply did not happen (or did not happen like he says) and the tradition does not exist. And it seems obvious to me that he’s not wrong. It is still open to us to go in a different direction, but this is the basis for the departure.