Podcast


Central Problem

Simon addresses the fundamental problem of complexity: how can complex systems exist? How can they evolve from simplicity? And how can we understand and describe them? The paradox is that the random evolution of complex systems seems statistically impossible given the available time — yet such systems exist.

The answer lies in hierarchical structure: complex systems do not assemble all at once but through stable intermediate subsystems. This hierarchical architecture not only explains the evolution of complexity but also determines the dynamic properties of systems and the very possibility of their description and comprehension.

The problem has implications spanning biology, physics, social sciences, organization theory, and artificial intelligence — all fields to which Simon contributed decisively.

Main Thesis

Simon‘s central thesis is that hierarchy is the architecture of complexity. Complex systems are almost universally hierarchical because hierarchical structure is the only one that can evolve in reasonable time and can be understood by finite minds.

The Parable of the Watchmakers (Hora and Tempus): Two watchmakers build watches of 1000 parts. Tempus assembles each watch as a single unit: if interrupted, he loses all his work. Hora builds stable subassemblies of 10 parts, then assembles them into groups of 10, then these into the final watch. With random interruptions (p=0.01), Hora is about 4000 times more productive than Tempus. The principle: stable intermediate forms exponentially accelerate evolution.

Near Decomposability: Hierarchical systems have a crucial property: intra-component interactions are much stronger than inter-component interactions. This means that:

  • In the short run, each subsystem behaves almost independently of the others
  • In the long run, subsystems interact only in an aggregate manner

The house example: rooms divided into cubicles, walls with different insulating capacity. Temperature equilibrates first within each room, then between rooms.

Describing Complexity: Hierarchical structure makes possible the economical description of complex systems. System redundancy can be captured recursively. The world is “nearly empty” — most things interact weakly with most other things.

Historical Context

The essay was presented in 1962 to the American Philosophical Society, at a crucial moment for complexity science. Simon — Nobel Prize winner in economics (1978) and pioneer of artificial intelligence — was developing a unified theory connecting his work on bounded rationality, organization theory, and problem solving.

The intellectual context includes: Wiener‘s cybernetics, Bertalanffy‘s general systems theory, Shannon‘s information theory, and early artificial intelligence research at Carnegie Institute of Technology. Simon sought common principles that would cut across these disciplines.

The essay also implicitly responds to Jacobson‘s speculations about the thermodynamic improbability of biological evolution, showing how hierarchy resolves the temporal paradox.

Philosophical Lineage

flowchart TD
    Wiener[Wiener] --> Cibernetica[Cibernetica]
    Shannon[Shannon] --> Informazione[Teoria informazione]
    Cibernetica --> Simon[Simon]
    Informazione --> Simon
    Bertalanffy[Bertalanffy] --> Sistemi[Teoria dei sistemi]
    Sistemi --> Simon
    Darwin[Darwin] --> Evoluzione[Teoria evoluzione]
    Evoluzione --> Simon
    Simon --> Complessità[Scienze della complessità]
    Simon --> IA[Intelligenza Artificiale]
    Simon --> Organizzazione[Teoria organizzazioni]

    class Wiener,Shannon,Bertalanffy,Darwin,Simon internal-link;

Key Thinkers

ThinkerDatesMovementMain WorkCore Concept
Simon1916-2001Cognitive ScienceThe Architecture of ComplexityHierarchy, near decomposability
Wiener1894-1964CyberneticsCyberneticsFeedback, control
Bertalanffy1901-1972Systems TheoryGeneral System TheoryOpen systems, equifinality
Shannon1916-2001Information TheoryMathematical Theory of CommunicationEntropy, information
Darwin1809-1882Evolutionary BiologyOrigin of SpeciesNatural selection

Key Concepts

ConceptDefinitionRelated to
HierarchySystem composed of interrelated subsystems, each in turn hierarchic down to the elementary levelSimon, Systems Theory
Near DecomposabilityProperty whereby intra-component interactions are much stronger than inter-component interactionsSimon, Complexity
Stable Intermediate FormsSubassemblies that persist long enough to serve as building blocks for more complex assembliesSimon, Evolution
State DescriptionDescription of a system in terms of its static properties (blueprint)Simon, Epistemology
Process DescriptionDescription of a system as a sequence of operations that generate it (recipe)Simon, Epistemology
SpanNumber of immediate subsystems in a hierarchic systemSimon, Organization

Authors Comparison

ThemeSimonvon BertalanffyWiener
Central focusArchitecture of complexityGeneral properties of systemsControl and communication
Key principleHierarchy and decomposabilityEquifinality, open systemsNegative feedback
ApproachAnalytical-empiricalTheoretical-generalMathematical-engineering
EvolutionCentral (stable forms)SecondaryMarginal
ApplicationsOrganizations, AI, biologyBiology, sociologyEngineering, neuroscience

Influences & Connections

Summary Formulas

  • Simon: Complex systems are almost universally hierarchical because hierarchy is the only architecture that can evolve in reasonable time and can be understood by finite minds.
  • Parable of Hora and Tempus: The presence of stable intermediate forms reduces evolution time from exponential to logarithmic in the number of elements.
  • Near Decomposability: High-frequency dynamics (internal to components) + low-frequency dynamics (between components) = comprehensibility and economical description.
  • Redundancy: Simple description of complexity is possible because hierarchical systems are highly redundant — few types of subsystems in various combinations.

Timeline

YearEvent
1948Wiener pubblica Cybernetics
1955Jacobson stima tempo evoluzione (troppo lungo senza gerarchia)
1956Simon, Newell, Shaw creano Logic Theorist (prima IA)
1957Simon lavora su organizzazioni e razionalità limitata
1962Simon presenta “The Architecture of Complexity”
1969Simon pubblica The Sciences of the Artificial
1978Simon riceve Nobel per economia (razionalità limitata)

Notable Quotes

“Among possible complex forms, hierarchies are the ones that have the time to evolve.” — Simon

“In a nearly decomposable system, the short-run behavior of each of the component subsystems is approximately independent of the short-run behavior of the other components.” — Simon

“If there are important systems in the world that are complex without being hierarchic, they may to a considerable extent escape our observation and our understanding.” — Simon