3 Later ideas

The English translation of The Logic of Scientific Discovery is palimpsestic: while translating it Popper intercalated comments, glosses, developments and corrections in new footnotes and appendices, as well as drafting a supplementary work, the three- volume Postscript of 1982–3. Opinion differs over whether all this is fully consistent. A case in point is ‘The Aim of Science’ section of the Postscript, published already in 1957, which argues that science aims at satisfactory explanations. It centres around a historical example (Galileo, Kepler, Newton), showing how each theory superseded and explained its predecessor. Satisfactory explanation, in addition to being testable, must fulfil other conditions, making it a rather stronger aim than falsifiability, one that may or may not be the same as the aim of science articulated at the end of The Logic of Scientific Discovery , of discovering ‘new, deeper and more general problems’.

Certainly Popper acknowledges some changes of view. Since metaphysical dispute surrounded the concept of truth, he carefully avoided using it in The Logic of Scientific Discovery, making do with logical relations (implication, tautology, contradiction). Later, convinced by Tarski’s work, he made free use of the concept of truth and of getting nearer to the truth (verisimilitude). Again, his criticisms of conventionalism in The Logic of Scientific Discovery were methodological. In later years they were also openly metaphysical, as Popper espoused a robust realism and indeterminism (1982b, 1983b ).

Throughout The Logic of Scientific Discovery there are Darwinian metaphors – the struggle for survival among theoretical systems, natural selection, fitness to survive – although, in the last pages of the book, the view that science is an instrument of biological adaptation is rejected. This Darwinian leitmotif became a controversial issue in Popper’s later work: did evolutionary biology yield to the same methodological analysis as physics? Were the central ideas of Darwin or of the modern synthesis falsifiable? To complicate matters, Popper changed his mind on this central question, viewing Darwinism as a historical hypothesis in Objective Knowledge, and as an unfalsifiable near tautology in Unended Quest. In contrast to his earlier view, Popper also began to advocate an evolutionary epistemology, that is, an attempt to explain the very existence of a truth-seeking science within the framework of natural selection, in effect to give a biological twist to Kant’s problem, ‘How is knowledge possible?’ His second Herbert Spencer Lecture (1975) treats both endosomatic and exosomatic adaptations as forms of knowledge. Biological considerations also weigh heavily in Popper’s part of The Self and Its Brain . Pitting himself against the reductionist materialism of most contemporary mind–body specialists, Popper there marshalled mainly indirect arguments for an interactive pluralism.

Reflections on biology seem to have been behind a bold new metaphysical initiative of 1967–8, especially the provocatively entitled ‘Epistemology Without a Knowing Subject’ (1972: ch. 3 ). Distinguishing the world of physical things from the world of mental things, Popper argued that objective knowledge is located in neither, but in ‘World 3’ – the world of humanly created objective contents of thought. Such intellectual products have an objective existence: theories, problems, problem- situations, theoretical situations and critical arguments have properties and logical interrelations that lack physical or mental analogues. Stored knowledge exists even if no living person retrieves it. Critics of World 3 find some of its consequences counterintuitive: for example, it contains not only all truths, but also all falsehoods, which thus have an equally objective ‘existence’.

In the late 1940s Popper published a particularly forceful and elegant system of natural deduction that is of considerable interest, both intrinsically and because in it he views deductive logic as the organon of criticism (1947a, 1947b). He reports that while he has repaired some defects in it, he never brought it to completion. His technical attentions became focused on the theory of probability, to which he had already contributed in The Logic of Scientific Discovery. The result was a highly abstract axiomatic system that made no explicit assumptions about any logical relations among the elements on which probability is defined, and thereby established that probability is a genuine generalization of deducibility. The system is open to many new interpretations of probability statements. A particularly important one, superseding the frequency interpretation of The Logic of Scientific Discovery, views probabilities as measures of the ‘propensities’ of states of the world to develop one way rather than another. In the Postscript (1982a) and A World of Propensities, this view was developed into a striking new metaphysics (see Probability, interpretations of §4 ).