Saturday, June 02, 2012

The Science of the Artificial

Harbert Simon, a professor of computer science and psychology as well as the 1978 Nobel laureate in Economics, pursued the “artificial science”, or knowledge about artificial objects and phenomena. According to him, artificial things have the following characteristics which distinguish themselves from natural:

- Artificial things are synthesized (though not always or usually with full forethought) by human beings
- Artificial things may imitate appearances in natural things while lacking in one or many respects, the reality of the latter
- Artificial things can be characterized in terms of functions, goals, adaptation
- Artificial things are often discussed, particularly when they are being designed, in terms of imperatives as well as descriptives

Human being as a simple system
Simon began with human beings. After studying aspects of imperfect problem solving activities of human beings and the other artifacts (i.e. computers), he claimed that human beings are quite simple. He said, “The apparent complexity of our behavior over time is largely a reflection of the complexity of the environment in which we find ourselves.” (page 53) What makes our behaviors complex is information of the situations full of variety. If we use metaphor to describe the problem solving activities of human beings, we can say as follows: human being is like a simple function having variables in it, and information is input to the variables.

There are some facts showing the very limited ability of human being. According to a study of Dansereau and Gregg, the times required for elementary arithmetic operations and for fixation of intermediate results account for only part of the total time for performing mental multiplications of four digits by two. Much of the remaining time appears to be devoted to retrieving numbers from the memory where they have been fixated. The fact suggests that the system of human beings is basically serial in its operation, i.e. we can solve the problems only through “one-by-one” style.

This simplicity of human being is counter-intuitive. However, Simon says to this notion that “Only human pride argues that the apparent intricacies of our path stem from a quite different source than the intricacy of the ant’s path.”

Science of designing artificial
Based on the studies of human beings’ behavior, Simon turns to the design of artificial. Although not comprehensive, he proposes the following evaluation criteria of artificial design:
1) Theory of evaluation: the artifacts act to maximize utility, or they could make decision based on statistic
2) Computational methods: (a) algorithms for choosing optimal alternatives such as linear programming computations, control theory, dynamic programming; (b) algorithms and heuristics for choosing satisfactory alternatives
3) The formal logic of design: imperative and declarative logics
4) The search for alternatives: (a) heuristic search; (b) allocation of resources for search
5) Theory of structure and design organization
6) Representation of design problems (the definition of the problems)

Curriculum for social design
He also discusses about the aspects we need to pay attention in social planning:
1) bounded rationality: the meaning of rationality can be quite complex if we think about the rationality in complex situations
2) Data for planning: the social system needs methods of forecasting, the use of prediction and feedback in control
3) Identifying the client: we need to think about for whom we create the system
4) Organizations in social design: an important goal of the design is to fashion and change social organization in general and individual organizations in particular
5) Time and space horizons
6) Designing without final goals: it means the system needs to be designed as an evolving system, full of flexibility for future uncertainties

It is outstanding to think about the system and human being in this way at the time when this book was published. The hardest task is to deny the complexity of human behaviors, and the out-of-the-box thinking enabled the author to come up with his idea.

Although the studies in this book seem very scattered to me, the study of unknown fields always tends to be like that, through out the history of the progress of theories: when the first thesis appears, it takes very complex methodology to solve problems, but as the theory evolves, there will be more formalized, simple and beautiful methodology.

Harbert A. Simon, “The Science of the Artificial” (3rd edition), The MIT Press, 1996/10/1)

No comments: