Shifting symbolic structures and changing theories
on the non-translatability and empirical comparability of incommensurable theories
pp. 125-148
in: Massimo Ferrari, Ion-Olimpiu Stamatescu (eds), Symbol and physical knowledge, Berlin, Springer, 2002Abstract
Scientific theories in the mature sciences do not consist of a collection of observational regularities, but are intended to capture what lies behind the phenomena. Theoretical concepts transcend the realm of immediately perceptible objects and processes; they are removed from the phenomenal world in that their referents typically remain hidden to the unaided senses. There are good reasons for conducting science by constructing theories rather than collecting observations. The history of science suggests that providing unified explanations of large areas of phenomena requires concepts that are detached from what can directly be experienced. A frequently mentioned example is taken from the history of thermodynamics. Drawing on observational quantities, one arrives at empirical regularities of the behavior of gases. Experience produces a number of lawful relations among quantities such as pressure, volume, and temperature. Adopting, however, a theoretical perspective and positing that the behavior of gases results from molecular collisions, governed by the laws of statistical mechanics, first, establishes connections among these relations, and, second, suggests corrections of them. Theoretical reasoning demonstrates that the empirically established regularities hold only approximatively, and moreover it indicates how the description is to be improved. Theoretical terms promote scientific understanding more effectively than observational concepts.