Herbert ("Herb") Alexander Simon

Herbert Alexander Simon was born in Milwaukee, Wisconsin on June 15, 1916, to Edna and Arthur Simon. Simon’s father worked for the Cutler-Hammer manufacturing company helping to design control devices. His father’s work was not a direct influence on Simon, but when he later began to study feedback-controlled devices, the connection to his father was a source of pride.

Simon’s family was quite ecumenical regarding religion: Simon himself was raised mostly as a Jew, though he attended a Lutheran Sunday school and later became a Unitarian, in part because there was “no room” in his personal theology for “any notion of a chosen people.” Simon’s highly analytic approach to religion was typical: he applied his own critical, rational faculties to everything—and everyone—he encountered. In particular, he applied it to the human mind and how it solved problems.

The human mind was central to all of Simon’s work, whether in political science, economics, psychology, or computer science. Indeed, to Simon, computer science was psychology by other means. Simon’s remarkable contributions to computer science flowed from his desire to make the computer an effective tool for simulating human problem-solving. To him, a computer program that solved a problem in a way that humans did not (or worse, could not) was not terribly interesting, even if it solved that problem far more efficiently than humans did. Conversely a computer program that failed to solve a problem might be a great achievement, so long as it failed in the ways that humans fail.

One of the most important outcomes of this approach to computer science was Simon’s development—and strong advocacy—of heuristic programming. Drawing on his studies of human psychology and of organizational decision-making, Simon noted that people intend to be rational but that they rarely, if ever, have access to all the information or all the time they would need to make the optimally rational choice. Thus, Simon concluded, we do not, because we cannot, solve problems by using exhaustive, precise algorithms. Rather, we must use simpler heuristics and accept satisfactory rather than optimal results in order to make decisions or solve problems. To use a common analogy: a safecracker with unlimited time can try every combination and thus can be assured of opening the safe eventually. The safecracker who operates in the real world, however, has limited time and so begins by trying combinations based on the owner’s family birthdays, anniversaries, and the like. This heuristic does not guarantee success, but it will work often, and when it works, gives results much more quickly.

Simon and his colleagues Allen Newell and J.C. Shaw employed this notion of heuristic problem-solving in the first successful AI program, the Logic Theorist (LT) of 1955-56, which was used to prove the theorems of Russell and Whitehead’s Principia Mathematica. In a wonderful ironic twist, Simon first used family members to simulate the workings of the Logic Theorist before it was programmed into a computer, so he had people simulate the workings of a machine designed to simulate the workings of people’s minds! Simon was so excited by LT that he famously announced to his undergraduate class the next semester that “Over the Christmas holiday, Allen Newell and I invented a machine that thinks.”

Significantly, in addition to employing principles of heuristic problem-solving, the Logic Theorist was an error-controlled, feedback “machine” that compared the goal state (the statement to prove) with the current state and performed one of a small set of basic operations in order to reduce the difference between the two states. The Logic Theorist was a remarkable success, and Simon, Newell, and Shaw elaborated on its basic principles in creating another renowned program, The General Problem Solver (GPS) in 1957-8. The GPS was not quite as universal as its name implied, but it was startlingly good at solving certain kinds of well-defined problems. Even more, GPS, like LT, appeared to solve them in much the same ways that humans did.

Simon’s novel approach to the computer was, in part, a product of his education at the University of Chicago in the 1930s, to which he won admission as an undergraduate by competitive exam. He flourished in the intellectual hothouse of interwar Chicago, attending few courses but reading widely—and debating fiercely—in political science, philosophy, and mathematics. The transition to graduate study at Chicago was nearly seamless for Simon, who relished the demanding, but unstructured, nature of work there in the Department of Political Science.

While at Chicago, Simon encountered the German philosopher, Rudolf Carnap, whose rigorous positivism meshed well with Simon’s emerging outlook. Simon also studied with the pioneering mathematical economist Henry Schultz, who introduced Simon to the burgeoning world of econometrics, to mathematical modeling, to sophisticated work on the theory of measurement, and to the Cowles Commission for Research in Economics, which was home at the time to eleven future Nobel Prize winners in Economics, including Simon. Simon believed that these mathematical economists were developing some powerful tools and techniques for modeling human behavior, but that they had an absurdly unrealistic image of the ability of humans to make rational choices. As he put it, “we need a less God-like, and more rat-like, picture of the chooser.” LT and GPS were intended to create just such “rat-like” models of how people actually solve problems in the real world.

Simon—with a series of collaborators—continued to develop programs designed to simulate the operations of the human information-processing system, ranging from programs that played chess, to the Elementary Perceiver and Memorizer (EPAM, co-created with Edward Feigenbaum) that simulated the processes of human sensory perception and learning, to BACON, which simulated the process of discovery in science. Throughout, he was a strong, even fierce, advocate of the computer program as the best formalism for psychological theories, holding that the program is the theory. The fullest statement of this belief was the monumental text, Human Problem Solving [7], authored by Simon and Newell in 1972, in which they introduced the notion of a program as a set of “production systems”, or “if-then” statements. The flip side of this coin was his insistence that computer simulation was an empirical science that taught us new and valuable things about ourselves and our world; simulation was not an exercise in elaborating tautologies.

Last, but not least, Simon believed that organization and structure were critical. Indeed, what his computer simulations simulated was not the actual physical operations of neurons in the brain, but rather the structure of problem-solving processes. The computer program thus could be a structural model of the mind in action, not a model of its specific physical make-up. Two of the key conclusions he drew about the structure of our human mental processes are that they are hierarchical and that they are associative. In other words, he believed that they have a “tree structure”, with each node/leaf linked to a branch above it. Significantly, to Simon, each “leaf” could either be one thing or a set of things—a list of things, to be precise, with the elements of a list possibly being other leafs with their own lists, and sub-lists. (Think of a “to do” list that contains the item “go grocery shopping”, an item that contains its own sub-list of items to purchase.) Since items on a list could “call” items on other lists, this model of the mind could work associatively within its basic hierarchic structure, creating webs of association amongst the branches of the mind’s tree.

To implement this hierarchical, associative model of the mind, Simon and Newell worked with Shaw (a programmer at RAND) to develop the first list processing language, IPL (Information Processing Language). While IPL, a low-level assembly language for list processing, was largely superseded by John McCarthy’s more powerful high-level list processing language LISP, it was a major influence on the development of later list-processing languages, including LISP itself.

As befits someone fascinated by organizations, Simon was an institution-builder as well as a researcher. In the world of computer science, his most significant institutional legacy is the world-renowned School of Computer Science at Carnegie-Mellon University. Simon, Newell, and their colleague Alan Perlis first created a Department of Computer Science in 1965, and they (and others) expanded it until it became its own school in 1988. In keeping with Simon’s interests in AI, simulation, software design, and human-computer interaction, the Carnegie-Mellon University School of Computer Science excels in those areas.

Simon died on February 9, 2001, having received not only the ACM Turing Award (shared with Newell in 1975), but also the Nobel Prize in Economics (1978), The National Medal of Science (1986), The American Psychological Association’s Lifetime Achievement Award (1993), the American Political Science Association’s Dwight Waldo Award (1995), and the Institute for Operations Research and Management Science Von Neumann Theory Prize (1988). He was survived by his wife of 63 years, Dorothea (who died in August 2002), and their children, Katherine, Peter, and Barbara.

Three useful sources of biographical information on Simon are:

1. Mie Augier and James March, Models of a Man: essays in honor of Herbert Simon, Cambridge, MA: MIT Press, 2004.A collection of essays by prominent scholars on Simon’s influence on them and their fields.
2. Hunter Crowther-Heyck, Herbert A. Simon: the bounds of reason in modern America, Baltimore, MD: Johns Hopkins University Press, 2005.The standard intellectual biography of Simon.
3. Herbert Simon, Models of My Life, NY: Basic Books, 1991. Simon’s own autobiography.

Author: Hunter Heyck