Diamond’s essay, “The Polywater Episode and the Appraisal of Theories,” is an example of the confrontational model of HPS, wherein history acts as evidence to test philosophical theories much in the way a scientific theory is tested. He uses the “polywater episode” to test a collection of Donovan, Laudan, and Laudan’s (D&Ls) theses about science (D&Ls 1988). In this essay, I discuss his history and philosophy of science as understood in the confrontational model. I find that the confrontational model encourages distinct historical and philosophical work, which pushes against an integrated HPS.
The polywater episode itself is short and entertaining. There was “anomalous water” with unique spectral properties. Scientists suggested this could be explained by either impurities in the sample or a new structure of water (Diamond 1988). In the end, it was impurities. Diamond takes the episode as (i) a scientific dispute with a clear victor, (ii) clearly delineated camps, and (iii) ample publication data (Diamond 1988). This allowed a thorough review of the experiments and actors involved in the episode.
Diamond does two interesting things in discussing the episode. First, he evaluates the D&Ls’ theses regarding a scientific failure. This is good: while successful scientific change is exciting, there is no prima facie reason why (some) theses should not apply to failures. Second, Diamond introduces some quantitative analysis in his historical approach. He performs an analysis of researchers' ages and their interpreted stance—whether for, against, or neutral toward polywater. This is a promising step forward, given that many of D&Ls’ theses are couched in quantitative terms but not appropriately treated as such (McAllister 2018).
It is important to note, however, that Diamond’s historical analysis is limited by the constraints of the volume and ultimately in service of testing D&Ls’ theses. This raises concerns about the kind and depth of historical research permitted by the confrontational approach. If the confrontational model encourages focusing on how historical events might test philosophical hypotheses, then good “confrontational history of science” locates and describes those historical events—and need not do much more.
Moving to the philosophical side, I found many issues with Diamond’s essay. I will focus on two issues, both of which I take to involve Diamond misreading the D&Ls’ theses.
Diamond’s treatment of T2.6 is philosophically hasty. He takes T2.6 to be “not consistent” with the polywater episode because “neither experiment could be interpreted as a crucial test” (Diamond 1988, 192). He thus seems to assume that if T2.6 is true of an episode, then that episode must be decided by a crucial test. But nowhere in T2.6 does it suggest that there must be a decisive test—only that the merits of the theory are determined by “a few” experiments (Diamond 1988).
Furthermore, in his treatment of GA4.5, he appears to misinterpret what guiding assumptions meant in the sense D&Ls intended. Guiding assumptions are theories that are “wide-ranging in application, and highly influential across a variety of scientific fields” (D&Ls 1988, 10). But it is unclear whether theories about polywater meet this criterion. It is plausible that polywater theory is a “specific, narrower” one that, while it may have major consequences if true, does not involve a “major assumption about the world” in the sense that Aristotelian and Newtonian physics do (D&Ls 1988, 10).
With these points in mind, we are reminded to be wary of misinterpretations despite “theory-neutral” language. While the theses were supposed to be couched and understood in neutral language, Diamond clearly took T2.6 to be synonymous with a crucial experiment thesis and GA4.5 to concern theories in general. He might attempt to establish his conclusions with more philosophical buttressing, but that was absent.
Let us take a step back for a moment. In answering, “How are the history and philosophy of science integrated, and is that integration successful?”, we might think it important to ask historians and philosophers. Philosophers would likely demand more conceptual clarity and philosophical argumentation. They might also express concerns about the theory-ladenness of both the historical data and the philosophical theory of science Diamond purports to test. Historians might express frustration with what appears to be a shallow and instrumental historical analysis of the polywater episode. They would likely be interested in extending Diamond’s historical analysis without restricting it to serve a philosopher’s theory of science. If both parties are dissatisfied, how might integration succeed?
Therefore, I suggest that the confrontational model raises distinct questions for philosophers and historians that push away from interdisciplinary integration. Furthermore, the historian’s lack of autonomy within the model may sow discord between the two communities.
To elaborate further, consider another important thesis often associated with the confrontational model: the division of historical and philosophical labor. This thesis holds that there are distinct roles for the historian and philosopher in HPS. Although not strictly implied by the confrontational model, the division of labor is friendly to it. In the confrontational model, history has a distinct role (i.e., providing data) from philosophy (i.e., articulating theory). Thus, there is a difference between the roles of philosophy and history within the model. And insofar as there are practicing historians, practicing philosophers, and very few integrated HPS departments, a division of human labor should be expected.
Having criticized the confrontational model, I cannot offer a positive proposal. Perhaps we should adopt a model in which the roles of philosophy and history are closely intertwined (e.g., Chang 2011). But we might also be skeptical that a united method would unite practicing historians and philosophers. I have no definitive answers—only the suggestion that reflection on the limitations of the confrontational model may offer further insight.
References
Chang, Hasok. “Beyond Case-Studies: History as Philosophy.” In Integrating History and Philosophy of Science, edited by Seymour Mauskopf and Tad Schmaltz, 109–124. Dordrecht: Springer, 2011.
Diamond, A. M. “The Polywater Episode and the Appraisal of Theories.” In Scrutinizing Science: Empirical Studies of Scientific Change, edited by Arthur Donovan, Larry Laudan, and Rachel Laudan, 173–210. Synthese Library 193. Dordrecht: Springer, 1988. https://doi.org/10.1007/978-94-009-2855-8_9.
Donovan, Arthur, Larry Laudan, and Rachel Laudan, eds. Scrutinizing Science: Empirical Studies of Scientific Change. Synthese Library 193. Dordrecht: Springer, 1988. https://doi.org/10.1007/978-94-009-2855-8.
McAllister, James W. “Using History as Evidence in Philosophy of Science: A Methodological Critique.” Journal of the Philosophy of History 12, no. 2 (2018): 239–258.