Incommensurability: do successive paradigms involve a meaning-shift that prevents point-by-point translation, and does that entail non-comparability?

Thomas Kuhn, answered from the texts and cited to the page.

The position in brief

Incommensurability is a feature of the relation between successive paradigms in a mature science: their key theoretical and observational concepts are not mutually translatable without loss, and the impossibility of translation is rooted in the holistic dependence of meaning on theoretical context.1 The F = ma derived from F = ma/√(1 - v²/c²) is not Newton's law; it just looks like it.

The Newtonian mass is conserved; the Einsteinian mass is interconvertible with energy; the two terms are the same orthographic mark with different meanings, embedded in different theoretical structures. The mature articulation is taxonomic: each paradigm carries with it a structured lexicon of kind terms that classifies the world into categories obeying the no-overlap principle — no individual object falls under two distinct kinds at the same level of the taxonomy.2

Across a revolution, the lexicon is restructured: terms are split, merged, demoted, promoted, or dropped, and the resulting taxonomies are incommensurable in the sense that there is no translation manual mapping one onto the other while preserving extension and intension together. Incommensurability is not non-comparability. The two sides of a square and its diagonal are incommensurable in the mathematical sense and obviously comparable.

Theories from different paradigms can be compared on which puzzles they solve, on the extent of their fit with phenomena, on their fruitfulness for new research; they cannot be compared by point-by-point translation of their consequences into a shared neutral language, because no such language exists.345

Sources

  1. Each of us was centrally concerned to show that the meanings of scientific terms and concepts— 'force' and 'mass', for example, or 'element' and 'compound'—often changed with the theory in which they were deployed. And each of us claimed that when such changes occurred, it was impossible to define all the terms of one theory in the vocabulary of the other.
    The Road Since Structure, pp. 41–42
  2. Second, it is a group (of scientists) that moves forward in time and an individual (the historian) who moves back, and the same descriptive terminology cannot uncritically be applied to both. An individual can, for example, experience a gestalt switch, but it is a category mistake to attribute an experience to a group.
    The Last Writings, pp. 101–102
  3. My concerns in this paper arise primarily from the second of these lines of argument, but the two are not independent, and I shall need also to speak of the first. With it I begin, attempting first to set aside some widespread misunderstanding of at least my own point of view. Even with misunderstanding eliminated, however, a damaging residue of the first line of criticism will remain.
    The Road Since Structure, pp. 42–43
  4. Two men who perceive the same situation differently but nevertheless employ the same vocabulary in its discussion must be using words differently. They speak, that is, from what I have called incommensurable viewpoints. How can they even hope to talk together much less to be persuasive. Even a preliminary answer to that question demands further specification of the nature of the difficulty.
    The Structure of Scientific Revolutions, pp. 160–161
  5. One aspect of that modification is particularly relevant here: I have in the past modeled the experience of scientists moving forward in time too closely on the experience of the historian trying to move back.
    The Last Writings, p. 101

Where the argument is made

  • The Structure of Scientific Revolutions chapters IX, X, XII
  • 'Reflections on My Critics' (1970)
  • 'Theory-Change as Structure-Change' (1976)
  • 'Commensurability, Comparability, Communicability' (1982, in The Road Since Structure)
  • 'Possible Worlds in History of Science' (1989)
  • 'Afterwords' (1993)
  • The Last Writings on the no-overlap principle and structured lexicons

How the argument was received

The position evolved between 1962 and 1992. The early statement in Structure IX-X was sketchy and rhetorically heavy; the late articulation in terms of taxonomic structure is more defensible technically. Both versions are in the corpus. The Kripke-Putnam externalist objection — that the causal theory of reference allows reference to remain stable across theoretical change, since reference is fixed by initial baptism and a causal-historical chain rather than by descriptive content — is the canonical philosophical objection and it has force at the level of natural-kind terms like gold and water. It does not touch the taxonomic-structure thesis: even when reference of individual terms is stable, the categorial restructuring that a revolution effects in the lexicon as a whole is not dissolved by externalist semantics. Bird's loose intensionalism is a proposal I have considered; it preserves something but loses the structural point.