Incommensurability of Paradigms

Core Idea

Kuhn argued that successive paradigms are partly incommensurable: they cannot be compared directly by neutral criteria because they operate with different conceptual frameworks and standards of evidence. What counts as a fact or good explanation differs between paradigms. This raises a puzzle for rationality: how can paradigm shifts be rational if paradigms cannot be evaluated by external standards? Kuhn argued that non-empirical factors (simplicity, elegance, fruitfulness) guide paradigm choice while preserving rationality.

Explainer

From your study of Kuhn's paradigm theory, you know that a paradigm is not just a theory — it is a comprehensive framework that includes exemplary problem solutions, shared standards of what counts as a good explanation, and criteria for what data are relevant. Incommensurability arises from this comprehensiveness: when a paradigm shift occurs, *the framework for evaluation itself changes*, not just the theory being evaluated.

The classic illustration is the concept of mass across the Newtonian and Einsteinian paradigms. Both use the word, but they mean different things by it. In Newtonian mechanics, mass is an absolute quantity independent of the observer; in special relativity, mass is invariant but energy-mass equivalence and the behavior of objects at high velocities transforms what "mass" does in the theoretical structure. You cannot simply translate Newtonian mass into Einsteinian terms and compare the theories term by term — the conceptual networks have different shapes. This is semantic incommensurability: key terms change meaning across paradigm shifts, so the theories cannot be straightforwardly compared by checking whether one predicts more than the other.

There is also methodological incommensurability: different paradigms prioritize different virtues and different kinds of data. Aristotelian physics found it important to explain why things have natural places; Newtonian physics treated this question as ill-formed. Ptolemaic astronomers could predict planetary positions at least as well as early Copernicans — the initial Copernican case rested not on predictive superiority but on simplicity and coherence. Since the standards of "simplicity" and "coherence" are themselves partly paradigm-relative, there is no fully neutral standpoint from which to adjudicate.

This creates the rationality problem. If paradigm choice cannot be made by applying paradigm-neutral standards, does that mean paradigm shifts are irrational — mere sociological events, conversions rather than arguments? Kuhn rejected full relativism. He argued that scientists do share some values — accuracy, consistency, scope, simplicity, fruitfulness — even if they weight them differently. Scientists can reason about paradigm choice; they just cannot reduce it to an algorithm. The shift is more like a gestalt switch than a deductive proof: once you see the new paradigm, the old problems dissolve and new ones become visible. Critics like Feyerabend pushed further toward relativism; defenders of scientific realism argue that paradigm shifts are cumulative enough to sustain claims of scientific progress toward truth.