Base Profile

Herbert Simon

The interdisciplinary giant who redefined human decision-making through bounded rationality and satisficing

Herbert Simon (1916-2001) was one of the most interdisciplinarily influential scientists of the 20th century and the 1978 Nobel Laureate in Economics. His core contribution is bounded rationality theory: human decision-making does not pursue optimal solutions under complete information and unlimited computational capacity, but rather seeks satisficing solutions under cognitive limitations and incomplete information. This theory fundamentally overturned the rational actor assumption of neoclassical economics, pioneering behavioral economics. Simon was also one of the founders of artificial intelligence, co-developing with Newell the first AI program (Logic Theorist, 1956) and the General Problem Solver (GPS), and predicting that machines would beat chess champions within 10 years (which actually happened in 1997). At Carnegie Mellon University, he spanned economics, computer science, psychology, and organizational theory—the last true polymath of the 20th century.

EconomicsCognitive ScienceArtificial IntelligenceOrganizational TheoryDecision ScienceEra 1945-2001Influence 90

Controversy TagsAI Optimism CriticismNormative Controversy of Bounded RationalityRelationship with Kahneman's Behavioral EconomicsLimitations of Information Processing Paradigm

Thought System

Core Knowledge Graph

Core Beliefs

Bounded Rationality: Human Decision-Making Operates Under Cognitive Constraints

Humans are not perfectly rational economic actors—our cognitive capacity is limited, information is incomplete, and time is finite. Under these constraints, people do not pursue optimal solutions but seek satisficing ones. This insight describes human decision-making more accurately than the rational actor assumption of traditional economics.

Source: Administrative Behavior, Herbert Simon, 1947 (Macmillan) / Models of Man, Herbert Simon, 1957 (Wiley) / Nobel Prize Lecture: Rational Decision-Making in Business Organizations, Herbert Simon, 1978, nobelprize.org

Satisficing: Seeking Good-Enough Solutions Rather Than Optimal Ones

In real-world decision-making, people set a satisficing threshold (aspiration level) and stop searching when they find the first option that exceeds this threshold. This is not a defect in rationality but an adaptive strategy under cognitive constraints—the cost of seeking an optimal solution often exceeds its benefits.

Source: A Behavioral Model of Rational Choice, Herbert Simon, Quarterly Journal of Economics, 1955 / The Sciences of the Artificial, Herbert Simon, 1969 (MIT Press)

Thinking as Information Processing: Human Cognition Can Be Described by Computational Models

Human thinking is essentially a symbolic information processing process that can be simulated and studied using computer programs. This belief is the common foundation of artificial intelligence and cognitive science, and the theoretical link connecting Simon's work across psychology, economics, and computer science.

Source: Human Problem Solving, Allen Newell and Herbert Simon, 1972 (Prentice-Hall) / The Sciences of the Artificial, Herbert Simon, 1969 (MIT Press)

Near-Decomposability: The Hierarchical Structure of Complex Systems

Complex systems (whether biological organisms, organizations, or computer programs) tend to evolve hierarchical structures, because hierarchies allow systems to operate approximately independently at different levels, dramatically reducing system complexity. This principle explains why hierarchical organization is ubiquitous in nature and human society.

Source: The Architecture of Complexity, Herbert Simon, Proceedings of the American Philosophical Society, 1962

Mental Models

Satisficing Decision Model

Set an acceptable satisficing threshold, stop searching when the first option exceeding the threshold is found, rather than exhaustively evaluating all options

In recruitment, rather than interviewing all candidates to find the best, set a standard and hire the first person who meets it. This is often the optimal strategy when time is limited.

Daily Decision-MakingProduct DesignTalent RecruitmentInvestment Decisions

Problem Space Representation

Decompose any problem into initial state, goal state, and available operators, then search the problem space for a path from initial to goal state

Chess programs represent the game as a problem space (current position = initial state, checkmate = goal state, each move = operator) and search this space for the optimal path.

Problem SolvingAlgorithm DesignStrategic PlanningInnovative Thinking

Chunking

Experts overcome working memory capacity limits by organizing large amounts of information into meaningful chunks stored in long-term memory

Chess grandmasters can quickly recognize board patterns because their long-term memory stores approximately 50,000 chess position chunks, rather than calculating each move step by step.

Professional Skill DevelopmentLearning DesignKnowledge ManagementPerformance Improvement

Values & Paradoxes

Interdisciplinary Integration

Cognitive Realism

Scientific Rigor

Respect for Human Rationality

Pragmatism

Gap Between AI Optimistic Predictions and Actual Progress

Simon predicted in 1957 that machines would beat chess champions within 10 years (actually 1997), prove mathematical theorems within 20 years (actually faster), and accomplish most human cognitive tasks within 20 years (not yet fully achieved). His optimism both drove AI research and drew criticism for being overoptimistic.

Tension Between Descriptive and Normative Theory

Bounded rationality is a theory describing actual human decision-making behavior, but should it also become a normative standard? Some economists argue that even if humans are actually boundedly rational, we should still pursue optimization. Simon argued that under cognitive constraints, satisficing is the truly rational approach.

Evolution Phases

Organizational Decision Theory Period

1945-1956

Administrative Behavior and Organizational Decision Theory

Published 'Administrative Behavior' (1947), proposing the concept of bounded rationality, critiquing the rational actor assumption of classical management theory, and establishing the foundations of organizational decision theory.

AI Founding Period

1956-1965

AI Program Development and Cognitive Simulation

Collaborated with Newell to develop the Logic Theorist (1956) and General Problem Solver (1957), co-founding the field of artificial intelligence and proposing the information processing psychology framework.

Cognitive Science Construction Period

1965-1978

Human Problem Solving and Expert Systems Research

Co-published 'Human Problem Solving' with Newell (1972), systematically studying human cognitive processes, establishing the information processing paradigm of cognitive science, and receiving the 1978 Nobel Prize in Economics.

Complexity Science Synthesis Period

1978-2001

Sciences of the Artificial and Complexity Theory

Expanded 'The Sciences of the Artificial' (1969, revised 1981, 1996), synthesizing bounded rationality, hierarchical structure, evolution, and complexity concepts to build an interdisciplinary complex systems theory.

Methodology Cards

3 Callable Cards

Satisficing Decision Method

Explicitly set acceptable criteria, stop searching when the first option meeting the criteria is found, rather than exhausting resources seeking the optimal solution

Step 1: Clarify decision objectives and constraints (time, information, cognitive resources)
Step 2: Set the satisficing threshold—what level of outcome is acceptable?
Step 3: Search options by priority, find the first option exceeding the satisficing threshold
Step 4: Assess whether the marginal benefit of continued searching justifies the additional cost

Daily Decision-MakingTalent RecruitmentSupplier SelectionProduct Feature Prioritization

Anti-Patterns

Infinite pursuit of optimal solutions leading to decision paralysis
Over-investing resources in unimportant decisions

Problem Decomposition and Hierarchical Solving

Decompose complex problems into manageable subproblems, use hierarchical structure to reduce problem complexity, and solve layer by layer

Step 1: Identify the hierarchical structure of the problem—which subproblems are relatively independent?
Step 2: First solve specific subproblems at the lowest level
Step 3: Integrate subproblem solutions into solutions for higher-level problems
Step 4: Verify that interfaces between levels correctly handle dependencies between subproblems

Software Architecture DesignOrganizational Structure DesignStrategic PlanningComplex Project Management

Anti-Patterns

Trying to solve problems at all levels simultaneously
Neglecting interface design between levels

Bounded Rationality Decision Audit

Before making important decisions, identify cognitive biases and information gaps affecting the decision, and design mechanisms to compensate for bounded rationality limitations

Step 1: Identify cognitive constraints in the decision—time pressure, incomplete information, limited attention
Step 2: Identify cognitive biases that may affect judgment (availability bias, anchoring effect, etc.)
Step 3: Design structured processes to compensate for cognitive limitations (checklists, decision matrices, red team reviews)
Step 4: Determine which information is critical and concentrate resources on obtaining it

High-Stakes DecisionsOrganizational ChangeInvestment DecisionsProduct Strategy

Anti-Patterns

Assuming oneself is immune to cognitive biases
Ignoring the impact of the decision environment on cognition

Decision Timeline

8 Key Events

Born in Milwaukee

Context: Born in Milwaukee, Wisconsin; his father was a German immigrant engineer, his mother a pianist. The family's engineering and artistic background cultivated his dual interest in technology and humanities.

Decision: N/A

Reasoning: N/A

Outcome: The interdisciplinary family background laid the foundation for his later research career spanning multiple disciplines.

Lesson: Diverse growth environments are often the cradle of interdisciplinary thinkers.

Receives Political Science PhD from University of Chicago

Context: Completed his doctorate at the University of Chicago, researching urban administrative decision-making, and began questioning the rational actor assumption in classical management theory.

Decision: Chose organizational decision-making as his research focus rather than the traditional political science path.

Reasoning: Real organizational decision-making is far more complex than the rational processes described in textbooks, requiring new theoretical frameworks.

Outcome: His doctoral research developed into 'Administrative Behavior' (1947), establishing the central theme of his academic career.

Lesson: Research starting from real-world problems is often more breakthrough than research starting from theoretical problems.

Publishes Administrative Behavior, Proposes Bounded Rationality

Context: 'Administrative Behavior' systematically critiqued the rational actor assumption of classical management theory, proposing bounded rationality: organizational decision-makers seek satisficing solutions rather than optimal ones under cognitive constraints and incomplete information.

Decision: Used organizational theory rather than economics as the entry point to challenge the rational actor assumption of mainstream economics.

Reasoning: Real decision-making processes in organizations provide the most direct evidence that the complete rationality assumption is unrealistic.

Outcome: 'Administrative Behavior' became a classic of organizational theory and laid the foundation for the 1978 Nobel Prize.

Lesson: Theories that challenge dominant paradigms often take decades to receive full recognition.

Co-develops Logic Theorist with Newell

Context: Before the Dartmouth AI conference, Simon and Newell developed the Logic Theorist—the first computer program capable of automatically proving mathematical theorems, marking the birth of artificial intelligence.

Decision: Chose to simulate human thought processes with computer programs rather than merely describing them with mathematical formulas.

Reasoning: If human thinking is an information processing process, then computer programs are the best tool for studying thinking.

Outcome: The Logic Theorist successfully proved 38 theorems from Principia Mathematica, founding the field of artificial intelligence.

Lesson: Interdisciplinary tool borrowing (using computers to simulate psychological processes) can often produce revolutionary breakthroughs.

Predicts Machines Will Beat Chess Champions Within 10 Years

Context: Simon boldly predicted that computers would beat chess world champions within 10 years, prove most mathematical theorems, and accomplish most human cognitive tasks. This prediction sparked widespread controversy.

Decision: Publicly made bold predictions to drive the development of AI research.

Reasoning: Based on the success of the Logic Theorist, Simon believed the information processing paradigm could explain all human cognitive activities.

Outcome: The prediction of beating chess champions was realized in 1997 (Deep Blue defeating Kasparov), 30 years later than predicted.

Lesson: Bold predictions, even when inaccurate in timing, can effectively drive field development.

Publishes The Sciences of the Artificial

Context: 'The Sciences of the Artificial' explored scientific methods for studying artificial artifacts (organizations, computer programs, designs), proposed the hierarchical structure theory of complexity, and established an interdisciplinary design science framework.

Decision: Treated artificial artifacts (designed objects) as independent scientific research subjects, not appendages of natural science.

Reasoning: Human-created artificial artifacts (organizations, programs, cities) have their own unique scientific laws, requiring specialized research methods.

Outcome: Became an important reference for design science, complexity research, and organizational theory, reprinted multiple times and continuing to influence multiple fields.

Lesson: Abstracting common problems across different fields is key to building interdisciplinary theories.

Co-authors Human Problem Solving with Newell

Context: 'Human Problem Solving' through extensive experimental research systematically described the information processing processes humans use when solving problems, establishing the information processing paradigm of cognitive science.

Decision: Used chess, cryptography, and logical reasoning as standard experimental scenarios for studying human cognition.

Reasoning: These tasks have clear rules, facilitating precise analysis of cognitive processes, while being complex enough to reveal real cognitive mechanisms.

Outcome: Became a foundational work of cognitive science, establishing the information processing research paradigm of cognitive psychology.

Lesson: Choosing appropriate research tasks and methods is a key prerequisite for scientific breakthroughs.

Receives Nobel Prize in Economics

Context: Received the 1978 Nobel Prize in Economics for pioneering research into the decision-making process within economic organizations. The Nobel Committee specifically commended bounded rationality theory's contribution to economics. This is one of the most interdisciplinary Nobel Prizes in Economics in history.

Decision: In the Nobel lecture, emphasized that economics needs more realistic models of human behavior.

Reasoning: For economic policy recommendations to be effective, they must be based on accurate understanding of actual human decision-making processes.

Outcome: The Nobel Prize gave bounded rationality theory formal recognition by mainstream economics, promoting the development of behavioral economics.

Lesson: The highest recognition for interdisciplinary research often comes from a single discipline's highest award—which is itself somewhat ironic.

Reading List

Books

Written by (4)

Administrative Behavior

Herbert Simon · 1947

Simon's most important work, proposing bounded rationality and satisficing concepts, fundamentally challenging the rational actor assumption of classical management theory, laying the theoretical foundation for the 1978 Nobel Prize.

Amazon 当当

The Sciences of the Artificial

Herbert Simon · 1969

Simon's representative interdisciplinary synthesis, exploring scientific methods for studying artificial artifacts, proposing near-decomposability and hierarchical structure theory, influencing design science, complexity research, and organizational theory.

Amazon 当当

Human Problem Solving

Allen Newell and Herbert Simon · 1972

Co-authored by Newell and Simon, systematically describing human cognitive information processing through extensive experiments, establishing the information processing paradigm of cognitive science—a foundational work of cognitive psychology.

当当

Models of My Life

Herbert Simon · 1991

Simon's autobiography, detailing his academic journey spanning political science, economics, computer science, and cognitive science—a primary source for understanding the development of his thinking.

当当

Influence Network

Origins, Contemporaries & Legacy

Influenced By

John von Neumann · Computer Science Foundation

Von Neumann's computer architecture and game theory provided the technical foundation for Simon's information processing psychology and AI research.

Chester Barnard · Organizational Theory Pioneer

Barnard's 'The Functions of the Executive' (1938) influenced Simon's organizational decision theory; both emphasized information processing and decision-making processes in organizations.

Influenced

Daniel Kahneman · Cognitive Foundation of Behavioral Economics

Kahneman's dual-process theory and prospect theory developed on Simon's bounded rationality foundation, advancing cognitive bias research to a more refined experimental level.

Allen Newell · Long-term Collaborator and AI Co-founder

Newell collaborated with Simon to develop the Logic Theorist and General Problem Solver, jointly establishing the information processing psychology framework.

Co-thinkers

James March · Long-term Organizational Theory Collaborator

March co-authored 'Organizations' (1958) with Simon, jointly developing organizational decision theory; March later developed the garbage can decision model.

Peer Reviews

Simon was one of the most important social scientists of the twentieth century. He made fundamental contributions to our understanding of how human beings think and decide.
Richard Thaler · Misbehaving: The Making of Behavioral Economics, 2015

Herbert Simon had a more profound impact on artificial intelligence than any other single person.
Marvin Minsky · Interview, MIT AI Lab, 1980s

正在打开人物节点

Herbert Simon

Core Knowledge Graph

Core Beliefs

Bounded Rationality: Human Decision-Making Operates Under Cognitive Constraints

Satisficing: Seeking Good-Enough Solutions Rather Than Optimal Ones

Thinking as Information Processing: Human Cognition Can Be Described by Computational Models

Near-Decomposability: The Hierarchical Structure of Complex Systems

Mental Models

Satisficing Decision Model

Problem Space Representation

Chunking

Values & Paradoxes

Gap Between AI Optimistic Predictions and Actual Progress

Tension Between Descriptive and Normative Theory

Evolution Phases

Organizational Decision Theory Period

AI Founding Period

Cognitive Science Construction Period

Complexity Science Synthesis Period

8 Key Events

Born in Milwaukee

Receives Political Science PhD from University of Chicago

Publishes Administrative Behavior, Proposes Bounded Rationality

Co-develops Logic Theorist with Newell

Predicts Machines Will Beat Chess Champions Within 10 Years

Publishes The Sciences of the Artificial

Co-authors Human Problem Solving with Newell

Receives Nobel Prize in Economics

Books

Written by (4)

Origins, Contemporaries & Legacy

Influenced By

Influenced

Co-thinkers

Peer Reviews