Allen Newell (1927-1992) was a foundational figure in both artificial intelligence and cognitive psychology. With Herbert Simon, he created some of the earliest AI programs — the Logic Theorist (1956) and the General Problem Solver (1957) — and developed the information processing theory of human cognition. His later work pursued the ambitious goal of unified theories of cognition — comprehensive theories that explain not just isolated phenomena but the full range of human cognitive capabilities within a single theoretical framework.
Key Structures
- Herbert Simon — A Nobel laureate and polymath who pioneered artificial intelligence and the study of bounded rationality — showing that human decision-making is rational within the limits of cognitive capacity.
- Problem Solving — The cognitive processes involved in finding solutions to novel, non-routine challenges — from well-defined puzzles to ill-defined real-world problems.
- Means-End Analysis — A general problem-solving strategy that works by identifying the difference between the current state and the goal state and selecting actions to reduce that difference.
Key Functions
- Co-created the first AI programs (Logic Theorist, General Problem Solver) with Herbert Simon.
- developed Soar cognitive architecture.
- proposed unified theories of cognition integrating all cognitive functions.
Key Contributions
The Logic Theorist (1956), presented at the Dartmouth Conference, proved mathematical theorems and is considered one of the first AI programs. The General Problem Solver (1957) implemented means-end analysis as a general problem-solving strategy, modeling how humans decompose problems into subgoals. Newell and Simon's Human Problem Solving (1972) presented detailed protocol analyses of human problem solving, establishing verbal protocol analysis as a rigorous research method and demonstrating that human problem solving could be modeled as search through a problem space.
In his 1990 book Unified Theories of Cognition, Newell argued that cognitive science needed to move beyond studying isolated phenomena and develop comprehensive architectures that account for the full range of human cognition. The SOAR (State, Operator, And Result) architecture, developed with John Laird and Paul Rosenbloom, implements this vision: a unified system that learns, solves problems, uses memory, and operates in real time, modeling hundreds of different cognitive tasks within a single framework. SOAR continues to be developed and has been applied to military training, intelligent tutoring, and human-robot interaction.