How to Discern a Physical Effect from Background Noise: The Discovery of Weak Neutral Currents

In this paper I try to shed some light on how one discerns a physical effect or phenomenon from experimental background ‘noise’. To this end I revisit the discovery of Weak Neutral Currents (WNC), which has been right at the centre of discussion of some of the most influential available literature on this issue. Bogen and Woodward (1988) have claimed that the phenomenon of WNC was inferred from the data without higher level physical theory explaining this phenomenon (here: the Weinberg-Salam model of electroweak interactions) being involved in this process. Mayo (1994, 1996), in a similar vein, holds that the discovery of WNC was made on the basis of some piecemeal statistical techniques—again without the Salam-Weinberg model (predicting and explaining WNC) being involved in the process. Both Bogen & Woodward and Mayo have tried to back up their claims by referring to the historical work about the discovery of WNC by Galison (1983, 1987). Galison’s presentation of the historical facts, which can be described as realist, has however been challenged by Pickering (1984, 1988, 1989), who has drawn sociological-relativist conclusions from this historical case. Pickering’s conclusions, in turn, have recently come under attack by Miller and Bullock (1994), who delivered a defence of Galison’s realist account. In this paper I consider all of these historical studies in order to evaluate the philosophical claims that have been made on the basis of them. I conclude that—contrary to Bogen & Woodward (1988) and Mayo (1994)—statistical methods and other experimental inference procedures from the “bottomup” (i.e. from the data to the phenomena) were insufficient for discerning WNC from their background noise. I also challenge Galison’s notion of the “end of experiments” and shall take the wind out of the sail of Miller and Bullock’s attack on some of Pickering’s claims, whilst rejecting Pickering’s sociological-relativist conclusions. Instead, I claim that an epistemic warrant from the ‘top down’ in the form of a theoretical postulate of the WeinbergSalam model was necessary for “ending the experiments”, i.e. for the acceptance of WNC as a genuine phenomenon in the scientific community.