Recombinant Architecture
WILLIAM MITCHELL

(Mitchel points at the large video-projection screen behind him) Being in this split condition is rather a nice way to introduce the issues I want to talk about this afternoon: the relationship between physical presence and telepresence and the relationship between physical space and digital space. Behind me is my image on a video screen. I feel a little bit like Citizen Kane: I wonder how many other speakers have felt this way?

I wonder what it would be like if I ducked behind the podium so you could only see the virtual presence on the screen behind me. What would be lost? Would you only lose a few rather redundant bits out of the presentation? I suspect not. There's something interesting about the relationship here between the physical body and the virtual body; the physical space of this auditorium and some other kinds of spaces being created here behind me. Roughly, that is the theme I want to develop as I talk to you this afternoon.

As homework for this discussion, I looked into the relationship of electronic media to the home--specifically, the history of speculation about interactive television in the home and its effect on domestic life. It turns out that these speculations go back surprisingly far.

This is the earliest one that I found. It has acquired some currency lately. This predecessor of the marketing videos of Time, Warner and AT&T is from a Punch cartoon of 1879. It shows something that the cartoonist labeled Edison's telephonoscope. It claimed to transmit light as well as sound. We see Edison's telephonoscope (imagined, of course) opening up a video window above the bedroom mantelpiece of a comfortable Victorian villa. The father and mother in comfortable English suburbia are teleconferencing with their children in Sri Lanka. Of course, this was vaporware, but perhaps no more than some of the services we're going to be promised soon. It just took a little longer than expected to get to market and that is a common enough occurrence.

Let's briefly examine this pay-for-view dejá vu. Firstly, we see a very big screen, viewed by a group from a distance. This is obviously a television and not a personal computer. The cartoonist was innocent of the concept of bandwidth, of course, so a lot of pixels are being pushed through the screen. We don't see any compression artifacts, either, which perhaps the cartoonist might have imagined if more was known about this technology in 1879.

The other thing that will gladden the hearts of the television people is that they're using nice simple remotes, not keyboards. These remotes are being used to capture sound. They're actually rather elegant. Both the father and mother are speaking into microphones. They're capturing sound, but we may imagine them to be performing some control functions as well. But where is it in domestic space? How does it fit into the idea of the Victorian home?

It's interesting to compare the Victorian and the modern conception of this. In very Victorian fashion, it's on the mantelpiece, like some kind of ornament. It's in the bedroom; the protagonists are in different time zones. You can see that it's afternoon in Sri Lanka and evening by the fire back in England.

Interestingly enough, this device is identified with the half, not a replacement. Somebody complained the other day that we have both here. The video screen above the half.

What this expresses is the powerful fascination of the idea that human kind can somehow overthrow the tyranny of distance. That we can maintain contact and perform transactions independently of bodily location. And we have heard that theme sounded over and over again during this conference. It is a powerful idea that will grip the human imagination for some time to come.

In the last three decades, this idea has moved towards reality. Before coming to the main part of my talk, I'd like to briefly recount a few of the highlights of the story of slightly more than three decades of translating this idea. It was imagined by the Victorians, as were many scientific ideas before they came to fruition. I'd like to take you through some of the landmarks along the way to making this a reality. First, in the sixties, we began to see the first working prototypes of interactive computer graphics systems.

This is a very famous one that had tremendous impact on the way designers thought about interaction with machines. This is Ivan Sutherland in 1963, working on a system called sketch pad, which introduced many of the fundamental ideas of interaction still with us today.

By 1967, we start to see the appearance of head tracking devices coupled with stereo displays and the first explorations of the idea of immersive virtual reality.

1969 was another landmark that I remember vividly: ground zero point for development of the ArpaNet, which later grew into the Internet. Very unobtrusively, at UCLA on the west side of Los Angeles, just a little bit more than twenty five years ago, the ArpaNet went into action. It started to pump out bits. I was a young assistant professor at UCLA and I remember that we gave the people who installed that first mode a lot of grief about the idea of a one-mode network. But, pretty soon, there were two and three modes.

Interestingly enough, the ArpaNet began as an instrument of the military. As I recall, it was an instrument of social purpose right from the start. What mainly went on in those very early days was exchange of E mail for social purposes: greetings with friends, making dates remotely and so on.

By the early 1990's, we had begun to see the emergence of commercial products that gave a combination of relatively inexpensive computation and audio/video transmission. So here's a typical example of a work station, but that box doubles as television set if it's set up differently. What we see here on the screen is finally a low-cost, easily proliferated realization of the kind of thing the Victorians imagined. We see distant people on the screen, as well as an open computer-aided design window. People are discussing a project over the Net, the Internet, in this case. And there is a video camera on top of the work station screen. A configuration very much like that slide I showed you. We are looking at remote interaction combined with computation.

The other thing that emerged in the early 1990s was effective cyberspace navigation tools. The World Wide Web had been around for a while and was a bit stodgy, frankly. It finally met Mosaic, which was really cute. It was a beautiful client-server romance and the rest is history. This created the capacity to move quickly to just the bits that you wanted, wherever they were on the Net. Over the last couple of years , we've seen an extraordinary proliferation of the use of this kind of technology to get around in the Web. So, from my point of view, we seem to be moving rapidly towards a worldwide information infrastructure that combines four things:

Firstly, the bandwidth and multi-media capability we associate with cable television--good video and audio.

Secondly, the switching capability we associate with the telephone system. You can rapidly switch and directly get from any point to any other in the Net.

The third part of the equation is the local storage and computing capability that we have come to associate with the personal computer.

The fourth part of the picture is the navigation capability demonstrated fairly recently by the World Wide Web and Mosaic combination. And it is being demonstrated in other ways by agent systems and so on. I think we can hope and fight for some additional characteristics that I believe are politically important and which certainly will not emerge automatically. Indeed, we'll be told that some or all of these things are too expensive and difficult. But let me enumerate some of the characteristics that do seem important to me. Other people have mentioned these.

Firstly, the importance of two-way symmetrical digital pipes in the plumbing of the information infrastructure. Otherwise, interactions are highly unbalanced and unsymmetrical. That seems to me a fundamentally undesirable political condition.

Secondly, something that has been said many times but bears repeating: some form of universal access to cyberspace. That is very difficult and begs the question of the details of precisely what that means. But it's something that becomes increasingly important as more and more services and economic opportunities are delivered digitally.

Thirdly, something that may be a little less obvious: the right of anybody to put a server on the Net and become a producer as well as a consumer of information.

So this is the kind of environment one sees emerging. And the interesting question from an architect's point of view is: what does this mean for the kinds of living spaces that we have? The kinds of domestic spaces we'll inhabit? The ways these domestic spaces and other kinds of spaces will come together to make community?

N e w . L i v i n g . S p a c e s

When I say community, I mean it in a broad sense: a physically definable or a virtual community. Or some combination and overlay of these things.

I think the fundamental implication from an architect's and an urban designer's point of view of the development of this sort of infrastructure is that digital information transmitted in large quantities through computer networks becomes a solvent that decomposes traditional architecture and urban patterns and allows the remaining fragments to recombine into new patterns.

It does this by breaking down the adjacency imperatives that traditionally have held the parts of the building together. Examples of this are plentiful. Let's take the very familar example of a main street bank. Look what's happened over the last couple of decades with the emergence of the automatic teller machine.

A bank building used to be a neoclassical edifice on the main street of a community. It not only accomodated all of the functions of the bank; it represented the power and prestige of that institution in the community. It was where you went when you wanted to conduct some transactions that involved money in significant quantities. Going to the bank was certainly a significant event in daily life.

Look at what's happened to banks now. The automatic teller machine came along and probably will go away as cash becomes even more dematerialized. But the condition that we had for some time was that the functions were accommodated in the bank building. And now they are being increasingly taken over by the network of automated teller machines and international money transfer software.

So that old architectual unity of the bank disappears. And then the fragments, the automated teller machines, end up in a lot of different locations. They end up combining with other building types where people need cash. You find them in supermarkets, student unions and transportation terminals. In South Central Los Angeles, you find them in police station lobbies because that's the safe place to pick up cash.

So a kind of recombination is taking place. I agree, it's not something to be commended, but that's the reality of the situation that has emerged.

You can go through building type after building type and see the beginnings and sometimes an advanced stage of this kind of decomposition and recombination.

I'm going to take you through a particular example of something we've been dealing with at MIT recently: the idea of a virtual design studio.

V i r t u a l . D e s i g n . S t u d i o

All architects and designers know that a design studio is a place where people come together in a well-lit environment to work on projects. This is a photograph of the design studio at MIT in the 19th century. You can see the architects at work. There are horizontal working surfaces where you can make drawings. There are places for physical models. The main thing going on in this bounded, coherent architectural space is discussion of the projects being designed. Design is largely a process of discussion, negotiation of ideas over the representations of what's being proposed.

This is the traditional kind of physical setting for that. But one can do this in a digital environment. We recently performed a very interesting experiment doing exactly this.

Here, represented by a simple screen image, is the kind of digital environment that will enable you to make a virtual design studio. This is a very simple way of doing it. What we have here is a Computer Aided Design window that gives access to the digital model of whatever it is that's being designed. This can be simple or complex access. The best thing is sophisticated application sharing on the software that maintains this model, but you can get away with some simpler things, too. And here we see the video windows giving access to remote participants in the discussion.

About a few months ago, we ran a design studio project that involved six universities scattered around the world. The Universities were: MIT, Washington University in St. Louis, Cornell University, University of British Columbia in Vancouver, the University of Hong Kong and the Technical University of Barcelona. So the participants were scattered around the world, across time zones and cultures--in North America, in Asia and in Europe.

This group of about fifty or sixty scattered participants, students and faculty members of different universities worked on a housing project for an area of Shanghai. That's an area of traditional courtyard housing. This is a computer-generated image of the massing of this housing and here's an oblique view of the traditional type of housing known as Leilong housing. It's a very beautiful type of housing fabric, but one that has certain problems in accomodating modern uses and is often in very poor condition.

In this project, site information and programmatic information was collected by the students in Hong Kong and mounted on a server in Hong Kong, where it was available to the students of all the other locations. Students at the different locations then used the computer-aided design systems available to them to develop design proposals. They also collaborated very intensively, exchanging the work that they produced by video conferencing through the Internet and by e-mail.

The interesting thing about this is that it worked very well. If you've ever tried this, you know doing video through the Internet is really terrible. It's really bad video: small, fuzzy, low-frame-rate images. But when setting up an environment with which you could discuss something that was on the virtual conference table before you, even this very poor telepresence did the job. It gave good insight into what would happen if you had a bit more bandwidth and some more powerful pumps to pump the bits through the pipes a little bit faster. So the students negotiated back and forth and all sorts of interesting things happened. Students collaborated with each other and criticized each other. It was a very interesting cross-cultural interaction.

The viewpoints of the students from Barcelona were very different from those of the students from Hong Kong or MIT. This created an extraordinarily rich discourse. Some interesting collaborative arrangements developed. Being very competitive, the MIT students and the Hong Kong students rapidly discovered that there was a twelve-hour time difference between the two locations. So the MIT students would work for twelve hours and then shoot the files across to Hong Kong. Then the Hong Kong students would work for twelve hours and send them back. In this way, they got a twenty-four-hour operation going across the time zones that allowed them to do twice as much work as anyone else. This also produced a long distance dating facility of some effectiveness. Okay between MIT and Cornell. Not so good between MIT and Hong Kong.

Anyway, the thing developed very effectively. This is an aerial map of Shanghai that was mounted on the server. It shows the area we were dealing with. A lot of this kind of visual information was available for students to work with. This is a street of Leilong housing.

The students produced and presented an extraordinary amount of work. It was critiqued by a virtual design jury, for which we did a synchronous hookup of the six different locations. The students presented their projects. They pinned them up inside their space and the critics from the six different locations made their comments. It was a very exciting jury. For those of you who are architects and have been on juries, it had the amazing property that you can shut somebody up just by closing their window. Many architects have wanted to do that. But one needs to handle these things in more subtle ways for them to really work.

We had various different technologies working, and C-U/C-Me, which was the lowest level technologically, is still very interesting. You can see windows open at half a dozen different locations.

I would forget that the students from Hong Kong were looking into my office continually, which was one of the interesting architectual conditions that began to develop. The most intriguing thing about this, as other people have discovered when they've done this sort of thing, is that a true sense of community really can develop even with such a crude situation as you see here. We'd simply leave the windows open and see our friends--and they did become our friends--coming and going in the different locations. And a genuine sense of space began to develop.

S o c i a l . a n d . C u l t u r a l . C o n s e q u e n c e s

I'd like to talk briefly about the social and cultural consequences of this sort of fragmentation and recombination of architectual space, which can take place as we introduce the circulation of bits. What I'd like to suggest to you is that we cannot begin to understand these implications by focusing solely on the logic of cyberspace as if it were an autonomous thing. We have to go beyond the very simplistic view that the infobahn will replace transportation with telecommunication and simply replace bodily presence with telepresence. That it will replace face-to-face meetings with disembodied transactions in cyberspace. What it's actually likely to produce is a considerably more subtle and complex redistribution of functions among buildings, transportation systems and computer networks.

Let me try to illustrate this to you with the homely parable of the pizza parlor, a building type that has gone through some transformations in the United States over the past fifty years or so. Not so long ago, pizza parlors were mostly to be found on Main street. They had advertising signs out front to pull in the customers. They had counters where orders were placed and cash was handed over. They had kitchens where the product was produced. They had spaces with tables and seating. All this was wrapped up in one small building. Customers walked in, made their orders at the counter and ate on the premises. It all worked very well within one particular architectural package. A coherent architectural package located in a particular way in the community.

In the era of the automobile that changed and a competing pattern emerged. The pizza parlor, perhaps by now a chain, didn't just rely on a sign. It advertised in the Yellow Pages and in the mass media. It moved from Main street to a location beside the highway. And it acquired a parking lot to accomodate the cars in which people now came. Many customers now phoned in their orders and had them delivered by car to their homes and workplaces.

This style of operation served a larger, more widely scattered group of customers. The road transportation and telecommunication systems began to play very significant roles in its workings. And its former architectural unity fragmented as consumption shifted from a single seating area to many different customer locations. And Main street began to die as the pizza parlor and other businesses left for more attractive sites. Soon, the old familiar Main street wasn't the place where you went to hang out anymore. The action was elsewhere.

Now, finally, somewhere in the mid-1990s something changed. Let's imagine we're looking back from about the year 2000. The street address became a network address. And the counter became a screen display that allowed the customer at any location to select toppings from a graphic menu, drop them on a virtual pizza, specify the size, see the price, and of course pay with some form of digital cash.

The kitchen transmuted into a nationwide collection of preparation points in locations carefully selected to provide maximum coverage of the market. When an order was received it was automatically routed to the preparation point nearest to the customer's address. The pizza was produced and packaged and it was delivered by radio-controlled vehicle. Now there was no street signs to attract clientele. There was no newspaper or television advertising either. Customers were attracted through storefronts in on line virtual malls and through network yellow pages listings. And there was no place left to eat in public.

Now the details of this story are a little bit fuzzy but I think we can anticipate that the new arrangements were a big hit. The pizza suppliers reached a much larger market. They could optimize their delivery and preparation operation. And more importantly, since customers consumed the product at home, they did not have to build and maintain restaurant facilities in expensive locations. They were out of the table business, waiters were a thing of the past. The customers liked it too, since they could get exactly what they wanted quickly, efficiently and reliably at a low price. But they sometimes missed the atmosphere of the old places. The conversations that developed there. And the opportunities that these places afforded to get out of the house. To meet old friends and make new ones. And to go to a place that made them feel that they were a part of a local community.

And of course, if you didn't have a home, even if you had some old fashioned non-digital coinage in you pocket, you couldn't get a pizza.

What's the conclusion of this? Cities of the near future will meet basic human needs through some combination of architectual settings, transportation systems for goods and our bodies and digi-telecommunications. Introduction of telecommunications doesn't just add new capability, it changes the whole system. It shifts and redistributes functions among various elements and sub-systems of the urban system. And it alters the relationships among the parts. There are potential gains in this new condition but also some important potential losses. Public space can erode. Changes in patterns of accessibility to services and economic opportunities can create winners and losers on a truly massive scale. Perhaps we'll have a situation where reinvigorated local communities with strong electronic connections to a wider world develop. You can imagine electronic Aspens or Palo Altos, Cambridges or Leidens. You can belong simultaneously to a small-scale, local, physically defined community and multiple-interest-defined virtual communities. Perhaps you can present yourself in different guises in different communities.

On the other hand, you can very easily see the development of a distopia consisting of fortified, highly serviced enclaves of domestic space for the priviledged surrounded by marginalized and disenfranchised inhabitants of discarded urban remnants. We could get a digital Jakarta.

The real challenge for architects and urban designers as these conditions change is to find those configurations of built space, transportation and telecommunications. Those combinations are hardware and software that will support the emergence of just, humane and creative communities.