There are many similarities between what we call symbolic systems and what is elsewhere called cognitive science. Both fields are interdisciplinary approaches to the study of mind in the abstract. Both fields seek a computational understanding of mind and intelligence. Here at Stanford, however, we have chosen the name Symbolic Systems rather than the more general term, cognitive science. This term, we believe, emphasizes what is unique about our approach to cognitive science.

First, the subject matter of symbolic systems does not begin with the study of agents who think, but begins with the wider class of agents that use symbols. The subject of the program is defined in terms of the notion of a symbolic system. Actually, the term symbolic system is ambiguous. It can mean either an abstract system of symbols, like a human or computer language, or it can mean an agent that uses such a system, either a human or a computer. This ambiguity is important, for we study both the abstract systems, and the agents that use them. The term ``symbolic systems'' emphasizes that computers are part of the subject matter proper, since they are symbolic agents. And even if it turns out to be impossible to create truly intelligent machines, computers will still be important examples of symbolic systems.

The second difference has to do with the methodology. In many cognitive science programs the starting point is the human mind and its states, independent of the world in which they are embedded. This approach, and the results it has achieved, are part of what a student will study in symbolic systems. But our program emphasizes another perspective as well--a perspective which a number of Stanford faculty hold: that the best theories of mind and intelligence will emerge by studying the relationship between agents and their environment. These researchers place emphasis on the external significance of language and thought--the relation between language and thought on the one hand, and what they are about on the other. While individual SSP faculty members differ in the degree to which they incorporate this claim into their research, all agree that as a methodological issue, it is of utmost importance and provides our program with a perspective not shared by many cognitive science programs.

These important issues are reflected in the program in concrete ways. Required coursework is designed to familiarize students with various types of symbolic systems including natural languages, computer languages and symbolic logic as well as getting students to work with and manipulate these systems. Students acquire an understanding of the philosophical and logical foundations of their subject. Empirical study of cognition is one aspect of the study of human agents. Philosophical analysis is another. So are mathematical models and their implementation in computer programs. The well-trained scientist in this field needs to understand all three approaches. Thus students in Symbolic Systems are required to learn empirical techniques in the study of cognition, absorb the philosophical foundations of their subject, and develop the mathematical skills needed to pursue the subject in a rigorous way.

Back: General Information
Next: What Can You Do With A Degree