# Herbert Simon The polymath who **broke the rational-actor model open from the inside**. Simon (1916–2001) was an economist, political scientist, cognitive scientist, and pioneer of artificial intelligence whose 1955 paper "A Behavioral Model of Rational Choice" introduced the concept of *bounded rationality* — the recognition that real decision-makers face computational, informational, and time constraints that make full optimization impossible. The 1978 Nobel Prize in Economic Sciences recognized this insight as foundational. Simon was also a co-founder of the cognitive revolution in psychology and the field of artificial intelligence, making him one of the few twentieth-century thinkers whose work shaped multiple disciplines. Simon mattered before Kahneman and Tversky and provided the conceptual scaffold that made their empirical program meaningful. If full rationality were possible, the heuristics-and-biases findings would be merely embarrassing; Simon's framework explained why they are *necessary*. ## Core Contributions ### Bounded Rationality (1955) The foundational claim: real decision-makers don't optimize — they *satisfice*. Faced with a problem, agents construct an aspiration level (a "good enough" threshold), search until they find an option meeting it, and stop. This is not a failure of rationality but a rational response to limited cognitive resources. The model accounts for actual human (and organizational) decision-making far better than the optimization model that classical economics assumed. The implications cascade through every social science: optimization models are normative ideals, not descriptive theories; institutions and routines exist to manage cognitive limits; what counts as "rational" depends on the cognitive equipment available. Bounded rationality is the intellectual root of behavioral economics, organizational theory, and the cognitive sciences. ### Administrative Behavior and Organizational Decision-Making Simon's 1947 *Administrative Behavior* (which originated his Nobel work) treated organizations as information-processing structures that compensate for individual cognitive limits. Routines, hierarchies, specializations, and procedures are not bureaucratic decoration; they are the machinery by which boundedly rational individuals collectively process more information than any of them could individually. Modern political-science accounts of bureaucracy descend from this framework. ### The Cognitive Revolution and AI With Allen Newell, Simon helped found cognitive science and artificial intelligence research. Their 1956 program "Logic Theorist" was the first AI program; their 1972 *Human Problem Solving* established the information-processing model of cognition that dominated psychology for decades. Simon's interdisciplinary range — moving fluently among economics, psychology, political science, and computer science — was unusual at the time and remains rare. ## KB Significance **1. Why optimization doesn't describe politics.** Naive political-science models assume voters and politicians optimize over policy preferences; Simon's framework explains why they don't. Voters satisfice on candidates; politicians satisfice on legislation; both rely on heuristics that produce systematic patterns rather than optimization. This is why political-marketing science works (it targets the heuristics) and why technocratic appeals to "what the optimal policy would be" frequently fail (the audience isn't operating in optimization mode). **2. Institutions as cognitive prosthetics.** The right's project of dismantling institutions — civil-service reform, agency capture, defunding of public administration — can be analyzed as the destruction of the cognitive prosthetics that make collective decision-making possible. Simon's framework makes visible what is being destroyed: not "bureaucracy" as a generic bad, but specifically the information-processing structures by which complex governance becomes feasible. Without them, decision-making collapses into either personalist whim or oligarchic capture. **3. The base for everyone after.** Kahneman and Tversky's heuristics catalog, [[Charlie Munger]]'s mental-models program, [[Gerd Gigerenzer]]'s ecological-rationality alternative, and most contemporary decision-science work all assume Simon's framework as the starting point. Disagreements within the field are downstream of Simon's foundational claim that rationality is bounded. ## See Also - [[Daniel Kahneman]] — Built the empirical heuristics-and-biases program on Simon's conceptual foundation - [[Gerd Gigerenzer]] — Took Simon's bounded-rationality framework in a different direction (ecological rationality) - [[Bounded Rationality]] — The full mental-model entry for Simon's foundational concept - [[Keith Stanovich]] — Extended bounded rationality with the capacity/disposition distinction - [[Decision Science]] — The hub where this work lives