Experimental Validity and Pragmatic Modes in Empirical Science

Abstract

The purpose of this paper is to show how the degree of experimental validity of scientific procedures is crucially involved in determining two typical pragmatic modes in science, namely, the preservation of useful procedures and the disposal of useless ideas. The term ‘pragmatic’ will here be used following Schurz’s characterisation of being internally pragmatic, as referring to that which proves useful for scientific or epistemic goals. The first part of the paper consists in a characterisation of the notion of experimental validity. The second part is focused on several historical examples illustrating how the above pragmatic modes relate to the question of experimental validity. The Michelson–Morley experiment is presented as a case of a highly valid and useful experiment preserved through the development of different theories (like, on the one hand, the ether theory, upheld by Fitzgerald and Lorentz, and, on the other hand, Einstein’s special theory of relativity). The concept of caloric will be discussed as an example of an idea that was discarded once it became useless, after heat was understood as a form of energy within the new frame provided by the kinetic theory of heat towards the middle of the 19th century.

Acknowledgements

This article was written while I was a Fulbright postdoctoral scholar at the Center for the Study of Language and Information, Stanford University, where I had the great fortune to be advised by Patrick Suppes. I would like to thank him for extremely useful discussions concerning the issues addressed here. I am also indebted to José Acacio de Barros for comments on a previous draft on the Michelson–Morley experiment, not to mention his patient explanations about various aspects of this case study. Thanks to both José Acacio de Barros and Gary Oas for showing me how ‘the real thing’ works in the lab. This paper benefited greatly from their demonstration on how the interferometer was actually used in that experiment. Finally, I would like to thank two anonymous referees of ISPSfor their helpful criticisms and suggestions.

Notes

[1] Steinle (1997) convincingly argues that there are different ways to engage in experimental practice, with theory–testing being just one of them. In particular, he distinguishes between what he calls ‘standard experimentation’ (focused on theory‐testing) and ‘exploratory experimentation’ (focused on formulating regularities). The central experimental procedure characteristic of the latter would be the systematic variation of experimental parameters, done in order to find out which of the various parameters influence a given effect, and which of them are essentially required to prompt such an effect. According to him, exploratory experimentation is specially useful in concept formation, since the attempt to determine regularities would often require the application of new categories. It goes without saying that Steinle’s distinction proves highly clarifying, and greatly contributes to reinforce some of the crucial intuitions underlying the account of experimental validity provided here. Another insightful discussion on how different observational procedures bring about different theoretical models is to be found in Galison (1997).

[2] Suppes (1998, 246–251) argues that the focus on practice as opposed to foundations is clearly recognizable in modern physics, and discusses the different, coexisting interpretations of quantum mechanics relying on the same experimental data. He also draws attention to other pragmatic aspects of science for which the present work provides some further historical evidence. First, the fact that ‘there is much broad agreement by both theoretical and experimental physicists on the truth or falsity of many kinds of observations made with or without refined instrumentation’ (Suppes 1998, 237), something that shows the special relevance of knowledge directly derived from practice. Second, ‘the pragmatic way in which physicists […] can use observations, computations and fragmentary theoretical models from many different viewpoints over many different centuries and take from the past work just that which is relevant and relatively sound’ (Suppes 1998, 238). The factor common to all these pragmatic moves is that they are aimed at focusing on those aspects that prove useful for developing whatever theory.

[3] According to classical test theory, predictive or concurrent validity (correlation between the predictor and the predicted) cannot exceed the square root of the correlation, between two versions of the same measure—that is, reliability limits validity (Pelham and Blanton 2003, 70–77; Carmines and Zeller 1979, 11–13).

[4] Although they added a new kind of validity, namely, content validity, which will not be discussed here, due to its minor relevance outside the social sciences.

[5] We should insist here on Mayo’s contribution to this topic within the field of philosophy and her emphasis on the significance of statistics for the epistemology of experiment.

[6] Shankland (1964, 28) explains that, before Michelson and Morley performed their most famous experiment, they had repeated Fizeau’s experiment with substantial improvements: ‘They found that the change in the observed velocity of light was accurately proportional to water speed and was altered by almost the exact amount predicted by the Fresnel formula’. After this remark, Shankland quotes a letter in which Michelson opposes this result to the one obtained in the MM experiment: ‘I also repeated the experiment with air with a negative result’.

[7] According to Shankland (1964, 17), ‘This famous optical‐interference experiment was devised to measure the motion of the earth through the aether of space by means of an extremely sensitive comparison of the velocity of light in two mutually perpendicular directions’.

[8] Swenson (1982, 42) reports, ‘Morley and Miller, who were more deeply committed to the analogy of optics with acoustics, carried on sporadically their attempts to perfect the aether‐drift apparatus. In Miller’s case especially the power of the acoustic analogy was overwhelming.’

[9] Shankland (1964, 16–17n1) quotes an article on ether by James Clerk Maxwell (Encyclopedia Britannica, 9th ed., vol. 8) where this difficulty is emphasised: ‘If it were possible to determine the velocity of light by observing the time it takes to travel between one station and another on the earth’s surface, we might, by comparing the observed velocities in opposite directions, determine the velocity of the aether with respect to these terrestrial stations. All methods, however, by which it is practicable to determine the velocity of light from terrestrial experiments depend on the measurement of the time required for the double journey from one station to the other and back again, and the increase of this time on account of the relative velocity of the aether equal to that of the earth in its orbit would be only about one hundred millionth part of the whole time of transmission, and would therefore be quite insensible.’

[10] As Shankland (1964, 19) points out, Michelson devised the form of interferometer employed in the experiment ‘for the express purpose of measuring an effect of the earth’s motion on the speed of light’.

[11] Shankland (1964, 33): ‘And at the 1900 International Congress of Physics in Paris, Kelvin had urged Morley and D. C. Miller (Michelson had left Case in 1889) to make another trial of the experiment, which they did with an even more conclusive null result than that obtained by Michelson and Morley.’ Laser versions of the MM experiment were also developed in the 1960s, producing outcomes that coincided with those of the original MM experiment.

[12] Swenson (1982, 42) vigorously emphasises the usefulness of the interferometer as follows: ‘Finally, I assert that the Michelson interferometer, both as an instrumental design and as a set of techniques for manipulating electromagnetic radiation, has proved itself perennially valuable. It therefore is as important as any theoretical construct to the history of modern science. It was conceived as an optical speedometer for the Earth, but it has become a tool for the measurement of megaparsecs as well as pico‐Angstroms. Should it not rank with the telescope, microscope, barometer and thermometer in the pantheon of instruments?’

[13] Swenson (1982, 42): ‘The Michelson‐Morley experiment was not a direct cause of Einstein’s work, “On the Electrodynamics of Moving Bodies” of 1905, but it was certainly an indirect influence on the young Einstein, because it had been a direct influence on Lorentz, Poincaré, Rayleigh, Kelvin, Drude, Mach and many other leading physicists. The postulate of the constancy of the velocity of light itself alone owed much to the notorious failure of the Michelson–Morley experiment. It owed still more, of course, to Michelson’s lifetime refinement of that value, essentially confirming its constancy.’

[14] Schwartz (1962, 64): ‘The common‐sense view that time and distance are absolute realities thus stands in direct contradiction to the fact of experience discovered by Michelson.’

[15] According to Psillos (1994), the achievements made during the development of the caloric theory were the following: the development of calorimetry (specific heat), the law of adiabatic expansion of gases, and Carnot’s theory of heat engines.