Popper's Proposal
The tough problem of verisimilitude, and its relevance for a realist conception of knowledge, was first seriously discussed by Karl Popper as an attempt to illuminate the rationality of the scientific enterprise.[68] In 1963, he gave the first formal explication of the verisimilitude notion, developing his theory, in Conjectures and Refutations, which he combined with his falsificationist epistemology elaborated in 1934 in Logik der Forschung.
The basic idea of his falsificationism was a Peircean fallibilist view, according to which all kinds of knowledge are uncertain and corrigible, so even our best physical theories are probably false. He thought, indeed, that scientific knowledge consists merely of conjectures, rather than certitudes, which scientists put to severe tests, whose results may or not refute these hypotheses, but never accept as true or as probable. By refuting some conjectures that have proved false, and substituting them with new ones not yet falsified, scientists come closer to the truth about the world insofar as scientific theories have a truth value depending on the relation between the theories and the extra-linguistic world. This realist faith, and his belief in the rationality and cognitively progressive character of the scientific enterprise, naturally necessitated a way to compare a new theory with its predecessor in order to evaluate its degree of improvement, namely its degree of verisimilitude, in this way substantiating the idea that science makes progress by replacing false theories with more truthlike (but probably still false) theories.Popper’s definition of verisimilitude was quite plausible. It stated, in non-technical terms: a theory is more verisimilar than another if the true consequences of the first include those of the second (namely, the former implies more true sentences than the latter), and if the false consequences of the second include those of the first (namely, the former implies fewer false sentences than the latter).
The truth approximation theory of Popper was demolished in 1974 by David Miller and Pavel Tichy, who independently demonstrated that, following Popper’s definition, a false theory can never be closer to the truth than another true or false theory. This result opened the way to the post-Popperian approaches to verisimilitude, all of which share the idea that the approximation to the truth should be the fundamental aim of scientific research. The concept of verisimilitude, therefore, has since then been developed by other authors, and today there can be distinguished at least five approaches (Kuipers 2000: Chap. 7), but none of them is without its difficulties.[69] So that, despite a good deal of work, such concept—whose importance actually goes well beyond both Popper falsificationism and particular kinds of realism[70]— seems to have eluded a precise, satisfactory characterization.We are here particularly interested in a very influential approach to the problem of progress in science, that of Evandro Agazzi,[71] which we are going to very briefly introduce in order to adopt it as a starting point for our reflections on the notion of verisimilitude.
3