InterCommunication No.12 1995
InterCreation
MAKOTO SEI WATANABE
INDUCTION CITY
In Search of a "Free Order": How to Guide, not Design, the City:
ALA, the Ideal Architecture for Diversity
The Computer as a Thinking-support Tool
No particular "method" may be necessary when it comes to designing architecture. The decisive answers arrive by intuition in the course of considering the specified requirements, such as floor space and function. They take shape as naturally as a smooth take-off after a good approach run, and as unpredictably as a revelation of the divine. Good intuition, the scientists often say, is superior to logic.
In dealing with the city, however, that is not the case at all. The city is an aggregate of architecture beyond your direct control. No matter how excellent an architect you may be, you cannot design the architecture on all the adjacent sites. Even so, if you resign yourself to doing nothing vis-*-vis the surroundings, nothing will change. With the city far from desirable as it is now, there needs to be a method by which to deal with the "unmanipulatable" elements. My research originates in the desire to indentify such a method.
You can make a conceptional drawing of how architecture will look at its completion, but not of a modern city. This is because a city is a living, constantly moving entity, and in its present it is always in the process of moving. What can be done is to focus not on the "hardware," i.e., compositional elements of the city, but on its "software," or mechanisms of movement. In that sense, we are not so much designing as guiding the growth of the city.

ALA, the Ideal Architecture for Diversity

Ants are often referred to as an example of living organisms other than human beings that build colonies. Small though it is, an ant is a complex and sensitive creature. Even with the latest technology, we cannot reproduce an ant. Little is known, in fact, about such basic questions as how ants swarm about sugar scattered on the ground.
It is possible, though, to predict their behavior with the help of computer. If a group of codes are input for the dots of light in the computer and the dots let loose on the display surface, they may move about like so many ants swarming around split sugar. Both are totally different in shape, yet if they show the same behavioral pattern, we may infer that real ants behave according to a combination of simple codes like the computer dots. If, then, you let the computer ants loose, for example in a cold area--felt "cold" by the computer--and observe their behavior, you may be able to tell something about the behavior of real ants in winter.
Even without understanding all the physical attributes of an ant, its behavior is thus predictable to a certain extent. That which is predictable is controllable. This is how the simulation test works. The computer-programmed dots of light represent artificial "life." Creating artificial life is not like creating robots. Any shape whatsoever is acceptable. The crucial thing is that many dots ("living beings") programmed with simple codes freely build colonies, decrease and increase in number, and "evolve" by incorporating the programs of others, thus following a diverse behavioral pattern like that of real living creatures.
This is not to say that evolution of the city can be inferred from observing the lives of ants, but simply that living creatures and the city are similar in that both are composed of numerous elements which, despite each of them being governed by relatively simple rules, are related to one another in such a complex manner that the evolution of the whole is hard to predict. In other words, we can define what happens in the world of human beings as "the city."
Entities with attributes such as these are generally called complex systems. The validity of that term aside, there are many other "systems" with similar attributes. The same principle applies to the weather, to economics, and to the ecologies of other animals.
New methods are presented here from that approach. We may not know all the laws that govern complex cities, but what if we try to simulate the diverse configurations of the city by use of partially-applicable codes. . . With this in mind, we devised what we call ALA (artificial life architecture).

The Computer as a Thinking-support Tool

One of our other objectives is to make the most of the computer as an aid to the thinking process. Computers are used mainly in architectural design today, for computer-aided design (CAD) and computer graphics (CG), mainly as a manual tool in draftsmanship.
This role is of course important, but computers are capable of much more. They can be used not only to rationalize and make more efficient the tasks we used to do completely by hand, but to accomplish things only a computer can do. Like enhancing the quality of architectural-design involvement in the city. In that pursuit we sought to develop a "thinking-support" program.
What is presented here are first-stage models of this program concept. They are: (1) life-simulation of an entity of diverse elements through the combination of simple, partial codes, and (2) mental simulation that visualizes abstract concepts and situations and promotes intuitive understanding.
Both are significant for meaning on the abstract level and for their potential as application tools. This study presents not specific results but new methodologies.

Direction: Makoto Sei Watanabe (FAX:81(JAPAN)-3(TOKYO)-3829-3837
ARCHITECTS' OFFICE, Lecturer, Yokoyama National University
Graduate School and Kyoto Seika University)
Assistance: Yasuhiro Nakano (Yokohama National University Graduate
School)
Studies: "City of the Sun God," Yuko Oto;
"The Self-and Other-Determining City, Shugo Ishii;
"The City of Correlative Wave Motion," Yasuhiro Nakano;
"The City of Distorted Spaces," Junko Akamatsu;
"The Comparative District City," Kohei Takara;
"The Momentarily Materialized City," Takakuni Yukawa.
(All of the above are graduate students at Yokoham National University.)
Support: Koji Tamiya (Architects Office)

Note: All of the research published here is the product of original computer
programming. No existing packaged software was used.
GO TOP
Title Page
No.0 Contents
 Internet Edition Contents
Magazines & Books Page