RESUMO
In this paper, we distinguish Quine's thesis of holism from the related Duhem-Quine problem. We discuss the construal of holism which claims that the effect of falsification is felt on a conjunction of hypotheses. The Duhem-Quine problem claims that there is no principled way of knowing how falsification affects individual conjuncts. This latter claim relies on holism and an additional commitment to the hypothetico-deductive model of theory confirmation such that it need not arise in non-deductive accounts. While existing personalist Bayesian treatments of the problem make this point by assuming values of priors for the conjuncts, we arrive at the same conclusion without invoking such assumptions. Our discussion focuses on the falsification of equiprobable conjuncts and highlights the role played by their alternatives in ascertaining their relative disconfirmation. The equiprobability of conjuncts is discussed alongside a historical case study.