Technical Rationality

In the AI Story

The framework Schon attacks has an institutional elegance that accounts for its persistence. Universities teach the theory. Professional schools teach the technique. Practitioners apply both. The sequence feels natural — of course you learn the science before you practice the art. Medical students study anatomy before touching patients. Law students master constitutional theory before drafting contracts. Engineering students work through differential equations before designing bridges. The curricular architecture assumes that competent practice is deductive application, and that the hardest part of professional education is transmitting the formal knowledge on which application depends.

Schon's objection cuts deeper than a preference for hands-on learning. He argues that technical rationality systematically misidentifies what professionals do when they are at their best. The formal knowledge is real and necessary, but it addresses a different question than the one practice actually poses. The theory tells you what to do once you know what problem you are solving. Practice demands that you first figure out what problem you are solving — the operation Schon called problem setting — and no theory covers that figuring out. The science does not tell you which science to apply. The technique does not specify which technique is appropriate.

The intellectual symmetry between technical rationality and classical AI is not coincidental. Herbert Simon's Sciences of the Artificial (1969) framed all intelligent behavior as search through a problem space — an epistemology structurally identical to the one professional schools assume. Both treat intelligence as the application of formal knowledge to pre-defined problems. The computational theory of mind and the technical-rationality model are the same framework wearing different clothes. Schon debated Simon for decades. The machines built on Simon's epistemology have, paradoxically, vindicated Schon's critique: large language models work in ways that resist the clean hierarchy of basic science, applied science, and practice.

The AI moment converts Schon's critique from an epistemological argument into an existential emergency for professional education. When the articulable knowledge that technical rationality emphasizes can be reproduced by a system available for a monthly subscription, the professional premium must be grounded in something else. The credential reckoning is technical rationality's institutional reckoning. The schools that recognize this will restructure around the reflective practicum Schon proposed four decades ago. The ones that do not will produce graduates whose most expensive asset — years of articulable knowledge acquisition — is worth less than the tool their employers already pay for.

Origin

Schon developed the critique of technical rationality across two decades of research in professional settings — architecture studios, psychotherapy sessions, engineering firms, urban planning offices. His collaboration with Chris Argyris produced the framework that became canonical in The Reflective Practitioner (1983), which diagnosed a crisis of confidence in professional knowledge that was already visible in the environmental disasters, urban renewal catastrophes, and medical errors produced by confident, credentialed practitioners following the rules their training had taught them.

The crisis Schon identified was slow-moving in 1983, audible to those who listened but easy to ignore. The AI moment has converted the background hum into a siren. Every profession that depends heavily on articulable knowledge now faces the audit that Schon's framework predicted: the machines have separated, with unprecedented precision, what can be articulated from what cannot — and the part that cannot is the part that justifies the professional premium.

Key Ideas

Downward hierarchy. Theory flows to technique flows to application; the sequence is linear and the knowledge is formal.

High ground vs. swampy lowlands. Technical rationality works beautifully on well-defined problems and collapses where the problems must first be defined.

Problem solving vs. problem setting. The theory covers what to do when the problem is given; it says nothing about how the problem gets set.

Structural symmetry with classical AI. Simon's search-through-problem-space model and the professional-school curriculum assume the same epistemology — and the large language models built on that assumption paradoxically validate Schon's alternative.

The professional premium is under audit. When articulable knowledge commoditizes, only the tacit part — knowing-in-action, judgment, reframing — justifies the credential.

Debates & Critiques

The most sustained defense of technical rationality comes from fields where the high ground is genuinely wide — pharmacology, structural engineering, actuarial work — where formal knowledge does most of the job. Defenders argue that Schon overgeneralized from architecture and psychotherapy, domains where problem-setting dominates, to professions where it does not. The counter-argument is that even within apparently well-defined domains, the hardest and most consequential decisions involve judgments that no theory specifies. The AI moment has strengthened Schon's position empirically: the parts of professional work that machines replicate easily are precisely the parts technical rationality describes, and the parts that remain stubbornly human are the parts Schon spent his career defending.