Allen Newell's paper discusses the concept of physical symbol systems as a foundational contribution to cognitive science. He argues that the notion of a physical symbol system—a class of systems capable of having and manipulating symbols, realizable in the physical universe—is crucial for understanding the human mind. This concept, derived from computer science and artificial intelligence, posits that human symbolic behavior is a form of physical symbol manipulation. Newell emphasizes that this hypothesis is central to the search for a scientific theory of mind, akin to how the theory of evolution is central to biology. Newell outlines the constraints on mind, which include universality, symbolic behavior, development, learning, robustness, and real-time operation. These constraints guide the search for a system that can satisfy them, and the physical symbol system hypothesis is seen as a candidate that meets several of these constraints. He argues that the physical symbol system is a constructive definition of a class of systems that can be used to search for mind, and that it is essential to understand these constraints in order to advance cognitive science. Newell presents a paradigmatic symbol system, SS, which is a machine capable of manipulating symbols. SS is described as a system with memory, operators, control, input, and output. It is capable of performing various operations, including assigning symbols, copying expressions, writing expressions, reading symbols, sequencing actions, and conditional branching. The control of SS interprets the active expression and executes the appropriate operation, demonstrating the system's universality. Newell then discusses the universality of symbol systems, arguing that a universal machine can produce any input-output function. He addresses the challenges of defining universality, including the limitations of input and output domains, physical constraints, the decomposition of input into instruction and input-proper, and the problem of describing all possible functions. He concludes that the concept of a universal machine is relative to a class of machines and that the existence of such machines is a key aspect of the theory of computation. Newell demonstrates that SS is a universal machine by showing that it can simulate a Turing machine, a classic example of a universal machine. This simulation highlights the flexibility and power of physical symbol systems in modeling human symbolic behavior and cognitive processes. The paper concludes with the importance of understanding these concepts in the broader context of cognitive science and the search for a scientific theory of mind.Allen Newell's paper discusses the concept of physical symbol systems as a foundational contribution to cognitive science. He argues that the notion of a physical symbol system—a class of systems capable of having and manipulating symbols, realizable in the physical universe—is crucial for understanding the human mind. This concept, derived from computer science and artificial intelligence, posits that human symbolic behavior is a form of physical symbol manipulation. Newell emphasizes that this hypothesis is central to the search for a scientific theory of mind, akin to how the theory of evolution is central to biology. Newell outlines the constraints on mind, which include universality, symbolic behavior, development, learning, robustness, and real-time operation. These constraints guide the search for a system that can satisfy them, and the physical symbol system hypothesis is seen as a candidate that meets several of these constraints. He argues that the physical symbol system is a constructive definition of a class of systems that can be used to search for mind, and that it is essential to understand these constraints in order to advance cognitive science. Newell presents a paradigmatic symbol system, SS, which is a machine capable of manipulating symbols. SS is described as a system with memory, operators, control, input, and output. It is capable of performing various operations, including assigning symbols, copying expressions, writing expressions, reading symbols, sequencing actions, and conditional branching. The control of SS interprets the active expression and executes the appropriate operation, demonstrating the system's universality. Newell then discusses the universality of symbol systems, arguing that a universal machine can produce any input-output function. He addresses the challenges of defining universality, including the limitations of input and output domains, physical constraints, the decomposition of input into instruction and input-proper, and the problem of describing all possible functions. He concludes that the concept of a universal machine is relative to a class of machines and that the existence of such machines is a key aspect of the theory of computation. Newell demonstrates that SS is a universal machine by showing that it can simulate a Turing machine, a classic example of a universal machine. This simulation highlights the flexibility and power of physical symbol systems in modeling human symbolic behavior and cognitive processes. The paper concludes with the importance of understanding these concepts in the broader context of cognitive science and the search for a scientific theory of mind.