Matthew Lund - Can Frames Solve the Problem of Theory-Laden Observation?

The thesis that concepts are essential to scientific practice, understanding, innovation, and discovery, was a rather novel idea in philosophy of science when it was expressed, in 1958, by N.R. Hanson in Patterns of Discovery. While Hanson’s discussion, particularly where it dealt with gestalt shifts, had a decisive influence on Kuhn’s account of science, Kuhn’s picture of science came to eclipse that of Hanson. However, I wish to argue that when the two philosophers’ accounts of science are updated with contemporary accounts of conceptual structure, Hanson’s view provides a superior picture of science.
   Hanson argued that the integration of our perceptual experience by concepts not only aids us in rendering experience intelligible, but that this conceptualizing activity elicits expectations and incipient theoretical models, which can be tested and refined. Thus, for Hanson, theory-laden observation provides the first step in theory discovery and model refinement. Kuhn, on the other hand, emphasized normal science, and admitted to having been in throng to the Kantian idea that categories determine the content of experience. Kuhn’s account is mostly concerned with how science functions when all of the conceptual frames are in place. Kuhn is very quiet about the processes through which new paradigms (or frames) get created, and usually speaks of such processes more as acts of creation than of construction. Hanson, on the other hand, provides some rather careful histories of the development of revolutionary concepts. I reconstruct, using the frame model, Hanson’s account of Galileo’s gradual discovery of the law for free-falling bodies as a case study to show that Hanson’s model is capable of capturing the subtleties of frame creation and adaptation.
   A feature of Hanson’s philosophy of science that is deserving of attention is the distinction he draws between pattern statements and detail statements. Pattern (frame) statements make assertions about the conceptual ordering of an area of phenomena, while detail statements (attributes) make empirical claims on the assumption of a given pattern (Patterns of Discovery, pp. 87-88). Both types of statements are empirical, according to Hanson, since they answer to observation, but are tested and evaluated in different ways. Famously, Kuhn denied (at least early in his career) that pattern statements are testable. I provide a frame-based reconstruction of Hanson’s arguments concerning Kepler’s development of the elliptical orbit theory, and demonstrate that rational testing of pattern statements was essential to Kepler’s inferential procedure.
   I conclude by showing that the epistemic problems brought up by theoretical influences on observation can only be solved by having both an acceptable theory of concepts (like the frame theory) and a complete and testable model of scientific change. I argue that neither Kuhn nor Hanson possessed the former, but that Hanson came closest to having the latter.