Resisting the Historical Objections: The Discontinuity Strategy

Some scholars denied the cogency of the PMI: for Lewis (2001), Lange (2002) and Magnus and Callender (2004) it is based on inductive fallacies, like the NMA. Doppelt claimed that it is incoherent (2011: 310, 2013: 48-49, 2014: 282-283), but he seems to be wrong (Alai 2016: § 7).

The discontinuity strategy adds that since there are radical differences between past and present science, no inductive inference from the former to the latter is possible; therefore also (MMT₂) applies only to past science, and the NMA remains cogent for current science.

Hardin and Rosenberg (1982) claimed that many false theories cited by Laudan were not products of mature science, so from their falsity we cannot infer to the falsity of theories in mature science. Yet, there were many false theories even in mature science (Newton’s gravitation theory, Fresnel’s wave theory of light, Rutherford's model of the atom, etc.). Devitt (1984: 143-149) argued that progress in scientific methodology, observation instruments and experimental technology continuously improve the reliability of theoretical science, hence past failures should not be projected onto the present or the future. Still, even from the Antiquity to the XVIII Century scientific methodology, instruments, etc., had greatly improved, but all theories accepted in the XVIII Century have been subsequently rejected.

Gerald Doppelt (2007, 2011, 2013, 2014) argues that old superseded theories were radically false and their success cannot be explained by assuming that they were partly true, while current best theories are completely true and will never be superseded or corrected by better theories. However, this “is an illusion owing to our particular historical perspective” (Nickles, this volume), which prevents us to see that progress began already in the past, and it will go on in the future.

Nonetheless, the discontinuity strategy might be right that contemporary science is much more mature than Newton's theory (or any of the false theories cited by Laudan), exhibiting a greater degree of sophistication, unification and coherence, and that from the XVIII Century to now there have been much greater method­ological and technological improvements than from the antiquity to the XVIII Century. For instance, ether and phlogiston were never measured, while we can measure many properties of unobservable entities (Dorato 2007: 181).

Fahrbach (2011: 1283) argued that “current best theories enjoy far higher degrees of success than any of the successful but refuted theories of the past”. In fact “three-quarters of all scientific work ever done was done in the last 30-40 years”, and the exponential increase in the amount, diversity, and precision of scientific data and computing power marks a sharp difference between science today and only a few decades ago. Therefore present theories are incomparable to earlier ones as to the number and diversity of the tests successfully passed.

Still, it is questionable whether this difference between past and present science is radically qualitative or merely quantitative, and in the latter case it couldn't block the PMI: at most it might require some caution in the inductive extrapolation from past failures, because “successful science” is not a unitary kind, but rather a set of practices bearing family resemblances (Bird 2015, § 3); or it might suggest that current theories will be superseded in a much longer time than past ones. At any rate, it cannot block the MMT, because if past theories were false but successful, truth cannot be the only explanation of success, even for current theories. Therefore the discontinuity strategy needs in any case to be supplemented by the selective strategy.

8