# Incommensurability

> Rival frameworks may lack a complete common measure because their concepts, problems, or standards differ.

- HTML version: https://robbiepalmer.me/ideas/incommensurability
- Source: https://plato.stanford.edu/entries/incommensurability/

Two frameworks are incommensurable when they lack a complete common measure.
Thomas Kuhn and Paul Feyerabend used the term to challenge the idea that rival
scientific theories can always be translated into neutral language and judged
by permanent rules.

Incommensurability does not mean that comparison is impossible or that every
choice is arbitrary. It means comparison may require interpretation because the
sides do not hold every concept, problem, and standard fixed.

## Three sources of difficulty

**Standards can differ.** Rival communities may agree that accuracy,
simplicity, scope, and fruitfulness matter but weigh them differently. A choice
then cannot be produced by one rule that both sides apply in the same way.

**Observations can be organised differently.** Training and expectation affect
what a practitioner notices and how an observation is classified. This is one
consequence of [theory-ladenness](/ideas/theory-ladenness), not a claim that the
world has no part in constraining observation.

**Concepts can change together.** A term in one framework may have no exact
equivalent in another because its meaning depends on its relations to other
terms. Newtonian and relativistic uses of "mass" belong to different conceptual
structures. Replacing the word in each sentence does not complete the
translation.

Kuhn's later work concentrated on local, taxonomic incommensurability. The
problem concerns linked groups of kind terms whose classifications cross. It is
not a claim that whole languages are sealed off from one another.

## Comparison still requires work

Imagine that two schools assess reading progress. One groups readers by the
texts they can interpret without help. The other groups them by a standardised
score. Both can collect reliable observations, yet their categories answer
different questions. Ranking the groups on one scale would quietly adopt one
school's definition of progress.

Comparison remains possible. The schools can describe their procedures, study
how individual cases move between classifications, and ask which scheme fits a
particular purpose. What they cannot do is pretend that the category names alone
establish a shared measure.

This is why translation differs from substitution. A translator must learn how
terms function within each framework and may need examples, history, or a
bilingual practice that can show where the structures diverge. Some losses can
be stated precisely. Others become visible only when a case that was routine in
one scheme has no stable place in the other.

## Records across a change

Incommensurability matters whenever classifications or evaluation rules change.
An old record may remain legible while losing comparability with a new one. A
revised diagnostic manual, legal category, map legend, or scoring rubric can
alter what a label includes. Combining counts across that boundary without a
crosswalk creates a false time series.

A careful migration preserves the original values and their definitions,
records when the scheme changed, tests proposed translations against concrete
cases, and marks results that cannot be compared. Sometimes regeneration is the
only honest choice because the old observation omitted distinctions required by
the new framework.

[Paradigm shifts](/ideas/paradigm-shift) can create these boundaries by changing
a community's concepts and standards. [Postpositivism](/ideas/postpositivism)
turns the resulting difficulty into a discipline: state the framework, examine
its history and values, and avoid claiming a neutral comparison where none has
been built.

## Sources

* Oberheim, Eric, and Paul Hoyningen-Huene. ["The Incommensurability of
  Scientific
  Theories"](https://plato.stanford.edu/archives/fall2018/entries/incommensurability/). *The
  Stanford Encyclopedia of Philosophy*, 2018.
* Kuhn, Thomas S. *The Structure of Scientific Revolutions*. Fourth edition.
  University of Chicago Press, 2012.
* Kuhn, Thomas S. *The Road Since Structure*. University of Chicago Press,
  2000\.

## Related ideas

- [Bounded Context](https://robbiepalmer.me/ideas/bounded-context.md): An explicit boundary within which one domain model and its language have a consistent meaning.
- [Paradigm shift](https://robbiepalmer.me/ideas/paradigm-shift.md): A scientific community can replace the shared examples, standards, and concepts that organise its normal work.
- [Postpositivism](https://robbiepalmer.me/ideas/postpositivism.md): Empirical inquiry can pursue objectivity while accepting fallible evidence, theory-laden observation, and the influence of values.
- [Theory-ladenness](https://robbiepalmer.me/ideas/theory-ladenness.md): Observations depend on prior ideas about what to notice, how to measure it, and what the result means.

## Where it appears

- Blog post: [The Philosophy of Data Science](https://robbiepalmer.me/blog/2022-03-02-the-philosophy-of-data-science.md)

---

Markdown index of this site: https://robbiepalmer.me/llms.txt