Exemplars

Definition

Kuhn's account of how a scientific community coheres runs through concrete worked examples, not abstract rules: "paradigms may be prior to, more binding, and more complete than any set of rules for research that could be unequivocally abstracted from them." Practitioners converge instantly on how to handle a new problem while disagreeing, or simply having no answer, when asked to state the rule they're applying. In the 1969 Postscript, Kuhn splits the overloaded word "paradigm" into the disciplinary matrix (the community's whole shared constellation — symbolic generalizations, models, values, exemplars) and the exemplar itself, which he calls "the most novel and least understood aspect of this book."

In the Book

The case is pedagogical: physics students don't learn "force" from a definition; they learn it by grinding through the inclined plane, the pendulum, the orbit, until unfamiliar problems start looking like familiar ones. Ask working physicists to state the rules of their trade and you get hesitation and disagreement; give them a problem and they converge immediately — the knowledge lives in trained similarity-perception, not in criteria anyone could recite. Kuhn frames research traditions as Wittgensteinian family-resemblance networks, coherent through overlapping likeness among exemplars rather than through a shared essence, and credits Polanyi's notion of tacit knowledge for the underlying idea that practitioners routinely know more than they can state.

Why It Matters

This reverses a rules-first picture of expertise and practice transmission: onboarding, style guides, and playbooks fail when written as abstract rules alone, because what actually trains judgment is worked examples — model solutions, gold-standard cases, an apprenticeship of comparison. It also supplies a diagnostic: if colleagues recite the same definitions but diverge sharply on concrete cases, the shared practice is not as healthy as it looks; if they disagree on definitions but converge on cases, it usually is.