Gerry's Home Page Preliminary Materials Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Bibliography Appendix

Sec 11.3

11.3      Contributions to a System for Innovative Design

The effort to illustrate the functionality called for by the theory resulted in three major contributions to building computer support for innovative design: (a) a hypermedia knowledge representation substrate, incorporating: (b) a system of perspectives and (c) an end-user language. The design of each of these features has been thought through, both in terms of the functionality required by the theory and in terms of their usability in a practical computer system for design professionals. Each has also been prototyped in executable code and subjected to testing to confirm the implementability of the ideas. Various versions of these features, along with auxiliary functionality have also been incorporated in a series of design environments that have been shown to lunar habitat designers for feedback.

(a) The hypermedia substrate incorporates the power of the fine-grained hypertext in the original Phidias system, provides an efficient and scalable object-oriented database for persistence, incorporates multi-media nodes, and integrates the perspectives and language into the fundamental node and link structure. This hypermedia offers an extremely powerful and flexible knowledge representation system, whose control by the user is limited primarily by the lack of a fuller user interface. Adaptability by the user—or plasticity of representation—is critical according to the theory. The Hermes hypermedia contributes an example of a substrate for supporting such adaptability.

(b) The perspectives mechanism is a contribution to Computer Supported Cooperative Work. It allows individuals to organize their own versions of knowledge representations and to share them. This provides a tool for supporting the evolution of knowledge by starting with systematically organized domains and allowing users to inherit and modify these and to organize meaningful new domains. The virtual copying approach is an inherently efficient mechanism, which encourages consistency by eliminating unnecessary duplication of representations in multiple copies.

(c) The Hermes language is a contribution to end-user programming languages and programmable design environments. It suggests ways of reducing the programming doctrine that users have to learn or keep in mind. Much of the traditional programming language doctrine is suppressed by keeping the corresponding features tacit in the Hermes language. Also, the appearance of expressions in the language supports tacit understanding by making heavy use of user-defined domain terminology and by following several syntactic conventions of natural language. At the same time, when the computational structure of an expression must be made more explicit to be understood or modified, this can be done to some extent through interface displays and to some extent by exploratory execution. A programming language paradigm that was implicit in Phidias’ query language has been pushed forward, extended, and modified to the point of a powerful end-user language that can play key roles in a system to support interpretation.

 

Computer technology can contribute to human emancipation. By providing computationally active media of external memory, it can significantly extend cognitive capabilities within an increasingly complex world. However, that requires a people-centered approach in which machine computations are at the service of human judgments and interpretation. Mainstream software approaches have developed within a social context dominated by the interests of military, government, and multinational corporations, resulting in computer applications that replace people or that dictate how they think and work. This dissertation has tried to present design rationale to oppose the bureaucratic interests, a theory to guide people-centered software development, and example mechanisms for giving people innovative, shared control over software computations.

Go to top of this page

Return to Gerry Stahl's Home Page

Send email to Gerry.Stahl@drexel.edu

This page last modified on January 05, 2004