How does the physics community deal with the subsequent work of a scientist whose earlier work has been regarded as incorrect? An interesting case of this involves Joseph Weber whose claim to have observed gravitational waves was rejected by virtually all of the physics community, although Weber himself continued to defend his work until his death in 2000. In the course of this defense Weber made a startling suggestion regarding the scattering of neutrinos. I will summarize the history of gravity waves including the rejection of Weber's claim around 1975, his later work on gravity waves, and examine the reaction of the physics community to his neutrino hypothesis.

Calibration, the use of a surrogate signal to standardize an instrument, is an important strategy for the establishment of the validity of an experimental result. In this paper, I present several examples, typical of physics experiments, that illustrate the adequacy of the surrogate. In addition, I discuss several episodes in which the question of calibration is both difficult to answer and of paramount importance. These episodes include early attempts to detect gravity waves, the question of the existence of a 17–keV neutrino, and the question of the existence of a Fifth Force in gravity. I argue that in these more complex cases, the adequacy of calibration, in an extended sense, was both considered and established.

Experiments often disagree. How then can scientific knowledge be based on experimental evidence? In this paper I will examine four episodes from the history of recent physics: (1) the suggestion of a Fifth Force, a modification of Newton’s law of gravitation; (2) early attempts to detect gravitational radiation (gravity waves); (3) the claim that a 17-keV neutrino exists; and (4) experiments on atomic-parity violation and on the scattering of polarized electrons and their relation to the Weinberg-Salam unified theory of electroweak interactions. In each of these episodes discordant results were reported, and a consensus was later reached that one result—or set of results—was incorrect. I will examine the process of reaching that consensus. I will show that the decision was reached by reasoned discussion based on epistemological and methodological criteria. It then follows that we may use experimental evidence as the basis of scientific knowledge.

In this article I suggest a tripartite classification of scientific activity; discovery, pursuit, and justification. I believe that such a classification can give us a more adequate description of scientific practice, help illuminate the various roles that evidence plays in science, and may also help to partially resolve differences between “constructivist” and “epistemologist” views of science. I argue that although factors suggested by the constructivists such as career goals, professional interests, utility for future practice, and agreement with existing commitments do enter into pursuit, it is experimental evidence that is decisive in justification. I illustrate this with two case studies from the history of contemporary science, experiments on atomic parity violation and their relation to the Weinberg-Salam unified theory of electroweak interactions and the fifth force in gravity. I also answer some of the criticisms offered of my earlier account of the episode of atomic parity violation.