The Sciences of the Artificial
The Sciences of the Artificial is the 1969 book in which Simon argued that designed things — organizations, software systems, economic policies, curricula, institutions — deserve their own science, distinct from the natural sciences that study the world as it is. The argument rests on a single structural distinction: every artificial system has an inner environment (its design, architecture, internal logic) and an outer environment (the world in which it operates, the problems it addresses, the constraints it faces). The system succeeds when its inner environment is well-adapted to the demands of its outer environment; it fails when the two diverge. The book establishes that design, properly understood, is the rigorous management of this interface — a form of knowledge as demanding as any natural science, though one that modern universities had largely displaced from their curricula. The framework has proven remarkably durable and has acquired new relevance in the AI age, where the design of interaction structures between humans and AI systems constitutes the most consequential design problem of the era.
In the AI Story
The book went through three editions during Simon's lifetime (1969, 1981, 1996), with each edition expanding the framework's application to new domains. The 1969 edition focused primarily on organizations and decision-making. The 1981 edition extended the framework to include cognitive processes and the design of cognitive artifacts. The 1996 edition, Simon's final major statement, integrated his work on problem-solving, artificial intelligence, and complex systems into a mature theory of design as a distinct science.
The book's core structural insight — that every designed artifact operates at an interface between its inner logic and the outer world's demands — has proven broadly applicable. Engineers designing bridges must manage the interface between structural mechanics and the specific loads the bridge will bear. Administrators designing organizations must manage the interface between reporting structures and the market conditions the organization must navigate. Educators designing curricula must manage the interface between pedagogical content and the cognitive characteristics of the students who must learn it. In each case, the designer's task is not to build the inner environment in isolation but to manage its fit with the outer environment.
The book's most radical claim is that design deserves the same institutional investment as the natural sciences. Simon argued that engineering schools had increasingly displaced design teaching with physics instruction, business schools had replaced administrative judgment with economics, and computer science had replaced systems thinking with mathematics. The integrative, judgment-intensive art that had been the distinctive competence of the professional was being squeezed out by disciplines that studied the world as it is rather than how it might be. The argument has only become more urgent in the AI age, where the absence of rigorous training in design — particularly the design of human-AI interaction — has produced a generation of builders equipped to generate but not to evaluate.
Origin
Simon developed the book's arguments during the 1960s, drawing on his work across organizations, economics, psychology, and computer science. The unifying thread was his commitment to studying artificial systems — systems that exist because someone designed them — with the same rigor that the natural sciences applied to the world of natural law.
The 1969 edition was based on the Karl Taylor Compton Lectures Simon delivered at MIT in 1968. The lectures were explicitly designed to articulate a vision of design as a science worthy of institutional investment, addressed to an engineering school audience that Simon believed had been insufficiently attentive to the distinctive nature of design knowledge.
Key Ideas
Artificial systems have inner and outer environments. The system's design is the inner environment; the world it must serve is the outer environment; success depends on the fit between them.
Design is a science. The rigorous study of how artificial systems should be built, given the characteristics of the bounded agents who operate within them, deserves the same institutional status as the study of how natural systems are.
The interface is where design happens. The designer's task is not to optimize the inner environment in isolation but to manage its adaptation to the outer environment's demands.
Modern education has displaced design. The integrative, judgment-intensive art of design has been squeezed out of engineering, business, and computer science curricula by disciplines that study the world as it is.
The science of the artificial is more urgent than ever. The AI age has produced the most consequential design problem in the history of artificial systems: the interface between bounded human minds and unbounded artificial capabilities.
Appears in the Orange Pill Cycle
Further reading
- Simon, The Sciences of the Artificial (1969, 1981, 1996)
- Donald Schön, The Reflective Practitioner (1983)
- Harold Nelson and Erik Stolterman, The Design Way (2012)
- Richard Buchanan, 'Wicked Problems in Design Thinking' (1992)