Structural engineering and architectural form: are engineers playing 'catch up', enabling forms far removed from structural ideals? Asks Matthew Wells.

The capability of engineers to provide any progression or divergence in architecture is highly questionable. Contrary to David Dunster's suggestion in his Comment article introducing this issue's theme (p36), rationalist thinking in architecture does indeed tend to produce a series of architectural types that appear again and again. Where does a presumption of innovative progression come from? Is it not more like a continuous present, reworking a lexicon of forms only occasionally to new purpose? Ideals of steady improvement give way to the phenomenon seen on the ground, where an iconic prototype, bold in all its parts, degenerates into pale copies, a leaching away of detail and intent. Incidentally, all technical historians conclude that modern know-how could not engineer a trireme, a cathedral choir or the Hagia Sophia any better than was done at the time.

Moreover, the practice of architecture does indeed depend on the tensions between invention and repetition; the critical emphasis on originality rather than sheer fluency subverts this. Engineers don't seek form--they play catch-up. Our current methods come from the aerospace industry and the North Sea enterprise, the practice of enabling forms derived from concerns far removed from structural ideals. In parallel fashion, architects ask engineers to realise their own enquiries. (Not since the experiments of Michael Hopkins or Ian Ritchie have I seen an architect working rigorously through the potential of a material or a tectonic.)

My comment on the comparison David Dunster makes between Giedion and Benjamin is that the former relied on pictures, the latter words. The chief value in comparing the two is in uncovering the instabilities in the foundations of technical history, and the unease of its contemporary application. Giedion's connection of material and structural possibility with historical change in architecture appears to me contingent, only to be accepted in the absence of alternative explanations. Benjamin reconnects engineering to its own lyrical component in the face of others (Flaubert, Maupassant), who detected the monstrous in modernism. His contribution is only just returning to importance with new emphases on transparency, concealment, veiling and so on.

I hope Benjamin's historicism and sensitivities displace Giedion's activism. Benjamin placed engineering within a complete world-view of city and society, while Giedion was writing for other architects and only peripherally for engineers. The odd perspectives, crops and graphic character of his images de-centre them--little wonder he failed to detect a grammar of construction. (At the quantity surveyor's behest, we still fill steel frames with cut blocks then daub them.) The modernist inventory of ribbon windows and flat roofs are not engineer-derived; there is no intrinsic way to use the new artificial materials, steel and reinforced concrete, in the ways there are in respect of timber and masonry.

While architects, prompted by Le Corbusier, establish the promenade, engineers are somewhere else, articulating and abstracting forms: movement for us is always perpendicular to the Earth's surface and that is the big separation that will always remain. While acknowledging the claims made by Giedion/Benjamin in respect of the transporter bridge, and having built the 24th transporter bridge ever constructed (the only one in recent years), I would say their value is in their strangeness. They are marginalia rather than manifestations of a new relationship between architecture and engineering.

I love the trope of the engineer as architect's psychoanalyst, mediating on behalf of the public. I have often thought that structure should lie, as archaeology, under the revealed expression of a building. Psychoanalysis is, however, a debunked pathology--isn't structure best left where it is, often expressed, sometimes not even implied?

The third point of the Roberto Mangabeira Unger quotation, cited by Dunster, suggests that the problem in discussing style in relation to contemporary architecture lies in engineering and 'the increasing failure of physical constraints to determine the shape of buildings'. For me this has always been the axis on which the contemporary debate turns. Whatever else the constraints are, they are no longer structural: the baton has passed to the environmental engineer. More fundamentally, why is there always an assumption that engineering advance guided at least one stage of the Modern Movement? As historians have revealed the multiple strands in the so-called Enlightenment, are there not many 'modern movements', of which the technically driven or influenced one is simply a highlighted part?

The preservation of the open intellectual structure, the semantic images and the transparency referred to in Dunster's article are critical to the possibilities that engineering still has. Engineers should not be let off the hook by blaming surveyors for the schools our children are being led to and the hospitals we will duly be entering. It is incumbent on architects to demand more of their engineers, who will undoubtedly respond. They are capable of moulding their single discourse to contribute to the others cited in the Dunster proposition.

Matthew Wells is the author of Engineers--a Study of Structural Designers (London: Taylor & Francis), to be published this March.