underdetermination

Underdetermination is a situation in which the available evidence is insufficient to determine which of several competing theories or hypotheses is correct. In philosophy of science, it arises when multiple theories yield the same observable predictions, or when data can be reconciled with different theoretical packages by altering auxiliary assumptions. A central articulation is the Duhem–Quine thesis: empirical tests do not isolate a single hypothesis but test a network that includes the theory, background assumptions, and auxiliary hypotheses. Because the whole web can be adjusted, data can underdetermine which elements of the package are correct.

There are different emphases of underdetermination. Empirical underdetermination concerns inference from data to theory when no unique choice is forced. Semantic underdetermination concerns how terms or concepts are interpreted in the face of competing theories. In practice, underdetermination is thought to be common to some degree, though not necessarily total; new evidence, novel predictions, or the development of competing frameworks often narrows the field.

Responses to underdetermination typically appeal to non-empirical virtues—such as simplicity, coherence with established theory, explanatory power, and conservatism—along with formal approaches like Bayesian reasoning. Some philosophers argue that underdetermination does not undermine scientific rationality because inquiry proceeds by testing, revising, and expanding the theoretical network, gradually reducing ambiguity. Others emphasize the role of theory-ladenness in shaping what counts as evidence.