Duhem–Quine Thesis

Duhem's analysis of physical experiments

Duhem (1906) argued that what a physicist tests by experiment is not an isolated hypothesis but a whole theoretical group. When a prediction disagrees with observation, we learn only that at least one member of the group is wrong; logic does not say which one to change. He therefore argued that a crucial experiment, one that definitively eliminates one of two rival hypotheses, is impossible in physics, citing Foucault's experiment on the particle and wave theories of light. Duhem left such choices to the scientist's good sense (bon sens). Ariew (1984) stressed that Duhem's thesis was a relatively modest claim confined to physics.

Quine's holism and falsificationism

In "Two Dogmas of Empiricism" (1951), Quine attacked the analytic–synthetic distinction and reductionism, arguing that our knowledge is a web that meets experience only along its edges. Any statement can be held true come what may if we make drastic enough adjustments elsewhere in the web, and even laws of logic are in principle revisable, a strong form of holism. This sat uneasily with Popper's (1959) falsificationism, on which a single counterinstance should lead us to reject a theory. Popper responded with a methodological rule forbidding ad hoc auxiliary hypotheses. Grünbaum (1960) opposed the strong reading that crucial experiments are impossible in principle, and Harding (1976) collected the main contributions to the debate.

Underdetermination and later responses

The Duhem–Quine thesis fed into debates on the underdetermination of theory by evidence, the idea that the same evidence can equally support different theories. Kuhn (1962) explained why scientists do not readily abandon a paradigm in the face of anomalies, developing his theory of paradigm shifts. Lakatos (1970) proposed protecting a theory's hard core while adjusting a protective belt of auxiliary hypotheses, judging a research programme by whether it is progressive, that is, whether it yields novel predictions. Laudan (1990) criticized exaggerated claims that underdetermination undermines scientific rationality. Bayesians such as Howson and Urbach (2006) and Strevens (2001) showed how prior probabilities can apportion blame for a failed prediction between a hypothesis and its auxiliaries. Stanford (2006) raised unconceived alternatives as a new problem of underdetermination.

FAQ

What is the Duhem–Quine thesis in simple terms?

When an experiment contradicts a prediction, logic alone cannot tell whether the hypothesis under test is wrong or whether the instruments or some other assumption are. Hypotheses are always tested bundled with other assumptions.

Does the Duhem–Quine thesis mean science cannot be trusted?

No. It says that logic alone cannot decide which assumption to revise, not that scientific judgment is arbitrary. Approaches such as Lakatos's methodology of research programmes and Bayesianism offer rational criteria based on predictive success and accumulating evidence.

Sources reviewed: 2026-10-26

References

1. Duhem, P. (1906/1954). The Aim and Structure of Physical Theory (P. P. Wiener, Trans.). Princeton: Princeton University Press.
2. Quine, W. V. O. (1951). Two dogmas of empiricism. The Philosophical Review, 60(1), 20–43.
3. Popper, K. R. (1959). The Logic of Scientific Discovery. London: Hutchinson.
4. Grünbaum, A. (1960). The Duhemian argument. Philosophy of Science, 27(1), 75–87.
5. Kuhn, T. S. (1962). The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
6. Harding, S. G. (Ed.). (1976). Can Theories Be Refuted? Essays on the Duhem–Quine Thesis. Dordrecht: Reidel.
7. Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the Growth of Knowledge (pp. 91–196). Cambridge: Cambridge University Press.
8. Laudan, L. (1990). Demystifying underdetermination. In C. W. Savage (Ed.), Scientific Theories (pp. 267–297). Minneapolis: University of Minnesota Press.
9. Ariew, R. (1984). The Duhem thesis. The British Journal for the Philosophy of Science, 35(4), 313–325.
10. Howson, C., & Urbach, P. (2006). Scientific Reasoning: The Bayesian Approach (3rd ed.). Chicago: Open Court.
11. Stanford, P. K. (2006). Exceeding Our Grasp: Science, History, and the Problem of Unconceived Alternatives. Oxford: Oxford University Press.
12. Strevens, M. (2001). The Bayesian treatment of auxiliary hypotheses. The British Journal for the Philosophy of Science, 52(3), 515–537.