Feyerabend’s Realism …
P.K. Feyerabend Realism, Rationalism & Scientific Method: Philosophical Papers, Volume 1,CUP, 1981
Criticism is facilitated by proliferation: we do not work with a single theory, system of thought, institutional framework or until circumstances force us to modify it to or give it up; we use pluralities of theories … from the very beginning. The theories .. are used in their strongest form, not as schemes for the processing of events whose nature is determined by other considerations, but as accounts or determinants of this very nature.
[O]bjective reality has been found to be a metaphysical mistake.
[W]e decide to regard those things as real which play an important part in the kind of life we prefer.
Part 1: On the interpretation of scientific theories
4 Explanation, reduction and empiricism
6 Reply to criticism
Discussions of proliferation.
Part 2: Applications and criticisms
8 Introduction: proliferation and realism as methodological principles
Proliferation of views was recommended by John Stuart Mill [On Liberty] ‘on four different grounds.
Without criticism based on alternative views:
- One may reject a true view.
- Typically, one will not progress towards the truth.
- Even a view that happens actually to be true will be held dogmatically.
- One cannot understand the accepted view.
The rise, success and triumph of a new theory , point of view, or philosophy almost always lead to a considerable decrease of rationality or understanding.
Mills’ System of Logic is quoted, approvingly, as against Popper.
11 Realism and instrumentalism
According to realism [scientific] knowledge is descriptive of … features of the univers.According to instrumentalism even a theory that is wholly correct does not describe anything but serves as an instrument for the prediction of the facts that constitute its empirical content.
Feyerabend argues that realism is preferable to instrumentalism, because realistic theories are more testable. However, in some cases, such as wave mechanics, no realistic theory is possible.
12 A note on the problem of induction
[I]t is important to note that [the inductive] hypothesis is no longer concerned with the truth or even with the probability (in the objective sense) of the generalization whose use it commends. It asserts that it is reasonable to generalize a predicate that has been found to be instantiated in a finite number of cases. … Even the most careful thinkers are sometimes found to believe that proof of the modified hypothesis gives them the right to expect success, or success in the long run. [This has already been refuted.]
[Feyerabend shows that the above, weak, hypothesis] can indeed be refuted and that the procedure suggested by it can be shown to be undesirable.
[Suppose that we have a theory T that inductively entails F, and also] another theory T’ … which covers the facts supporting T , makes successful additional predictions A, and [inductively] entails F’ [inconsistent with F]. The test of the additional predictions may be regarded as an indirect proof that F’ and, thereby, as an indirect refutation of T.
Any [empirical] theory T under consideration … may be inconsistent with facts which are accessible only indirectly, with the help of an alternative T’. [Thus] it is reasonable to use not only [the given theory], but as many alternatives as possible.
16 Niels Bohr’s world view
This compares and contrasts Bohr and Popper.
2. Propensity: a part of complimentarity
Feyerabend notes that both Popper’s propensity and Bohr’s complimentarity differ from conventional probability theory in depending on the whole context, as in Jack Good’s theory.
8. Refutations of two objections
Feyerabend refutes Popper’s objections, in LSD, to Bohr’s theory. Of more general interest …
Only if it is assumed that these elements are systems which are in classically well-defined states can we derive Popper’s[objective] interpretation from the statistical character of the … theory … . However, it is also evident that this statistical character, taken by itself, is never sufficient for deriving such an assumption. … [F]rom the fact that a theory is statistical we can infer that it works with collectives [etc]. We cannot draw any inference about the individual properties of those objects … nor can we infer the category of the elements (i.e., whether they are for example objects, or transitions). Any such information would have to be given in addition to the laws characterizing the relationship between the frequencies ion the collectives.
10. Conclusion: back to Bohr!
Feyerabend is dismissive of Popper and – in some footnotes – very dismissive of Bunge.
He also has:
Problems of Empiricism
P.K. Feyerabend Problems of Empricism, Philosphical Papers Vol. 2, CUP 1981.
8 Consolations for the Specialist
7 A plea for Hedonism
[The] interplay between tenacity and proliferation … is … an essential feature of the actual development of science. … It may be the only means of eventing our species from stagnation. … Such an enterprise [‘mature’ science] is not only ill conceived and non-existent; its defence is also incompatible with a humanitarian outlook.
8 The role of reason in science
Is such a position irrational? … Yes, because there no longer exists a single set of rules that will guide us through all the twists and turns of the history of thought (science) … .
Briefly, regards the tendency for ‘normal’ science to pretend to be mature, with a single ‘true’ theory, as a consequence of the conservatism of most people who become scientists. He supposes that science would be better if it recognised and sought to explore the widest range of theories consistent with what was really known. It occurs to me that this may be true more widely.