Apart from shopping and "leisure," airports are being used ever more intensively by businesses. Virtual corporations - companies that have forsaken their headquarters for a life on the road - are big users of airports. The increased business conducted across national boundaries in the so-called "New Economy" has fueled demand for meeting rooms, exhibition and showroom facilities, business centers and other non-travel-specific facilities inside, next to, under, and on top of, most new airports. Airports have replaced green-field science and business parks as the epicenter of business real estate.

These multiple programs and agendas - operational, commercial, political - are one reason why writers have started talking about airports as cities of the air. Both cities and airports cover large areas; both are a complex of intersecting transport systems, economies, buildings, and people. But there is one crucial difference: cities have inhabitants. At airports everyone is transient. Herein lies one element of the airport's existential ambiguity.

A spectacular example of this trend is Kansai International Airport, which opened in September 1994. It's a $15 billion phenomenon described even by the can-do Japanese as "an exceptional endeavor for the country." The 500 hectare airport has a projected capacity for 160,000 aircraft movements a year, and is built on an artificial island about 5 km offshore from Osaka -- the most feasible option since the densely populated country is quite mountainous. The average water depth at the site is 18.5m, with unstable clay underneath. Preparation of the site began in 1987, and involved driving one million sand piles into the sea bed. They then built an 11km long sea wall, which outlined the island, and filled the resulting hole with 164 million cubic meters of soil. It must have taken a lot of buckets and spades.

Meanwhile, Renzo Piano, working with engineers Ove Arup and partners, won an international competition for the 300,000 m2 terminal. The central concept, the geometry of the main roof, was inspired by an airflow design and developed with reference to fractal geometry. Renzo Piano believes that an "air terminal should be structured as a diagram of the people moving through it. The entire building - structure, light and air movement - should complement the logic of passenger movement." To its credit, the airport has also segregated commercial activities on another level, allowing passengers more or less uninterrupted access from landside to airside. The 82.5m-long main beams over the 300m long main halls will afford travelers the pleasure of being part of a moving throng in a great space. The roof is airy and uncluttered and full of light. The design allows people to experience a drama of changing space, light and sights, familiarity intermixed, as the jury report put it, with "bold revelations of the future." Passengers arriving from the mainland will enter a great 'canyon' - the interface between nature and the machine - some 25m wide and 300m long with escalators, lifts and bridges slung between the four public halls.

At their first briefing meeting, Piano's team was presented with a 2,000 page summary of the technical briefing. As legend has it, the great man ceremoniously threw this $50 million document in the trashbin -- implying that he needed to start with, and hold onto, a clear design concept, and that the details could be worked out later. Amazingly, the design concept has survived a vast, fast, multi-billion dollar program more or less intact.

Kansai has successfully emerged as an architecturally coherent building. But the opportunity to spend $10 billion building a new island is not an option for most airport developers in the world. So the third design problem I want to discuss - the so-called cognitive dissonance that comes from being a rootless monad in an opaque system - is likely to remain pervasive in airports around the world unless new design strategies are developed.

Countless modern scholars, from Karl Marx to Baudelaire to Richard Sennett, have written about the alienation people feel in modern cities. Urban anxiety is part of our culture, almost familiar by now. But scientists have only recently addressed the phenomenon, having discovered the importance of what Hubert Dreyfus calls "situated understanding": the notion that a clear mental picture of the whole artificial environment contributes to mental health.

When it comes to giving us a "clear mental picture" of our environment, architecture has not exactly done a great job in recent decades. Design responses to the problem of complexity have tended to adopt one of three approaches:

1 - embrace it/celebrate it
2 - hide it
3 - make it familiar, reassuring, and homely

Airports have been struggling unsuccessfully to catch up ever since. Planners, besieged by data, normally think in terms of boxes. Occasionally someone has at least thought about system complexity when designing an airport: the result - for example at Roissy Charles de Gaulle - has often looked stunning on paper but been a nightmare to use. I must have passed through Roissy 30 times and I still have to ask where to catch the railway shuttle.

- +