Key Principle
Paradigms are transmitted not through explicit rules but through exemplars -- concrete problem-solutions (inclined plane, conical pendulum, Wheatstone bridge) that train scientists to perceive similarity relations between physically distinct situations. This is what Kuhn calls "the most novel and least understood aspect of this book" (Postscript). Scientists agree on which achievements are paradigmatic without agreeing on what abstract features make them so. The knowledge embedded in exemplars is tacit, communal, and non-rule-governed: it is "miscontrued if reconstructed in terms of rules that are first abstracted from exemplars and thereafter function in their stead" (Postscript). Paradigms are therefore logically and functionally prior to any rules extractable from them.
Why This Matters
This is the structural precondition for incommensurability. If paradigms could be decomposed into explicit rules, paradigm change could be modeled as a logical transition between rule-sets -- and revolutions would look like ordinary theory updates. Because paradigmatic knowledge resides in trained perception and practiced similarity-recognition, paradigm shifts require retraining at the level of doing, not merely updating beliefs. The exemplar mechanism also explains why scientists are expert practitioners yet "little better than laymen at characterizing the established bases of their field, its legitimate problems and methods" (Ch. V): competence resides in practice, not propositional knowledge. This transforms debates about scientific rationality: the demand that paradigm choice be governed by explicit rules misunderstands the nature of scientific knowledge itself.
Good Examples
- The Galileo-Huyghens-Bernoulli chain: Historical mechanics advanced by modeling new problem-solutions on previous ones under the single Principle of vis viva. Galileo saw a ball on an incline as like a pendulum; Huyghens decomposed a physical pendulum into Galilean point-pendula; Bernoulli modeled water efflux on Huyghens' pendulum. The verbal law remained constant -- what changed was the trained capacity to recognize actual descents and potential ascents in nature. Paradigm extension is driven by acquired similarity relations, not rule application (Postscript).
- The helium atom dispute: A physicist and chemist gave opposite answers to whether a single helium atom is a molecule -- the chemist said yes (kinetic-theory behavior), the physicist no (no molecular spectrum). This is not terminological confusion: "Their experience in problem-solving told them what a molecule must be" (Ch. V). Different exemplars for the same concept produced different ontological commitments.
- f = ma as law-sketch: Symbolic generalizations like f = ma are not fixed empirical laws but templates taking different concrete forms per situation (free fall, pendulum, harmonic oscillators, gyroscope). Understanding a textbook chapter does not enable solving its problems because the verbal content is "virtually impotent" without exemplar-trained pattern recognition (Postscript).
Counterpoints
- Tacit knowledge is not mystical. Against critics who charged Kuhn with making science rest on "unanalyzable individual intuitions," he replies: these intuitions are shared, group-tested, and acquired through training. They are analyzable -- but their analysis will be neurological, not rule-based. The knowledge is communal, not private (Postscript).
- The stimulus-sensation gap limits but does not eliminate objectivity. Stimuli are shared physical inputs (denying this is solipsism), but sensations are processed experiences that may differ because neural processing is partly conditioned by education. Color blindness going undetected until Dalton (1794) demonstrates that systematic perceptual differences can persist unnoticed -- same stimuli, different sensations. Yet "very few perceptual modes survive group use in a given environment," providing an evolutionary constraint on relativism (Postscript).
- Values provide rational guidance where rules cannot. The disciplinary matrix includes shared values (accuracy, simplicity, fruitfulness) that function as guides to theory choice. But values are "shared but variably applied," and this variability is functionally essential: it distributes risk across the community, ensuring some conserve while others innovate (Postscript).
Key Quotes
"Indeed, the existence of a paradigm need not even imply that any full set of rules exists." -- Thomas S. Kuhn, Chapter V
"The resultant ability to see a variety of situations as like each other... is, I think, the main thing a student acquires by doing exemplary problems." -- Thomas S. Kuhn, Postscript
"When I speak of knowledge embedded in shared exemplars, I am not referring to a mode of knowing that is less systematic or less analyzable than knowledge embedded in rules... Instead I have in mind a manner of knowing which is miscontrued if reconstructed in terms of rules that are first abstracted from exemplars and thereafter function in their stead." -- Thomas S. Kuhn, Postscript
"But our world is populated in the first instance not by stimuli but by the objects of our sensations, and these need not be the same, individual to individual or group to group." -- Thomas S. Kuhn, Postscript
Rules of Thumb
- When trying to understand how a scientific community works, look at the textbook problems students solve, not the abstract principles they recite. The exemplars reveal the paradigm more reliably than any explicit rule-set.
- If practitioners cannot articulate the rules governing their practice, this is not a sign of intellectual failure -- it is evidence that their knowledge is exemplar-based and tacit.
- Disagreements between specialists from adjacent fields (like the helium atom case) often signal different exemplar training, not factual error. Ask what problems each was trained to solve.
- Wittgensteinian family resemblance, not necessary-and-sufficient conditions, is how paradigm membership is recognized. Problems relate to each other through "overlapping and crisscross resemblances," not shared essential features.
- The perception vs. interpretation distinction matters: paradigm-dependent seeing cannot be corrected by better rules because it operates at the neural level, prior to deliberative interpretation.
Related References
- resolution-and-conversion - Why conversion requires retraining perception, not just updating beliefs
- progress-without-teleology - How puzzle-solving capacity (transmitted via exemplars) measures progress
- rules-of-thumb - Practical diagnostics built on the exemplar framework