| The celebrated philosopher-physicist Pierre Duhem appears to maintain virtually contradictory views. On the one hand, he claims that science does not aim to explain natural phenomena, where he assumes that an "explanation" strives to reveal the natural world underpinnings hiding "behind the veil" of observable phenomena. Despite these strong disavowals, he also insists that successful scientific theories should converge on "natural classifications" which allegedly provide "hints concerning the true affinities of things." But won't such relationships also lie "behind the veil"? These warring inclinations have created significant exegetical confusions, leading his interpreters to classify him as an antirealist, a realist and everything else in between. Duhem is clearly trying to get across some important methodological lesson about science. But what is it?;The trick is to align his philosophy more closely with the forms of physics he endorses. On this basis, I argue that Duhem's disavowals of "explanation" actually represent arguments against a dynamic laws picture of science: the doctrine that science must seek laws that track material systems according to the basic patterns of D-N explanation. He argues that many of natures most important hidden quantities (e.g., entropy and potential energy) were not discovered in a dynamical manner but were instead uncovered by stringing together relationships in the quasi-static manner employed in thermodynamics. Indeed, it is the deep relationships of the latter subject of the latter subject that supply paradigms of the "natural classifications" that Duhem seeks.;Once one follows through the details of his recommendations, employing concrete scientific examples, one realizes that Duhem's reflections on the scientific method greatly enlarge our appreciation of what the many varieties of "good science" can look like. This challenges many dogmatic presumptions about the scientific methodology that still prevail in contemporary philosophy of science. |
