Logical and Epistemological Approaches

A fundamental distinction in the approaches to verisimilitude one can choose, or better in the problems one wishes to resolve, needs to be underlined. Such distinc­tion was already clearly posed by Popper in his seminal work of 1963.

A preliminary definition of an appropriate notion of verisimilitude gives an answer to the logical problem, namely to the question about what we mean if we claim that one theory is more verisimilar than another. This problem tries to give an appropriate account of the concept, also determining its logical properties. The attempted solutions are various, but can be gathered together under three broad “families”; however, none of these approaches is unanimously considered as the best or most promising one (see Oddie 2014).

The epistemological problem of verisimilitude, instead, reads: how do we know that one theory is more verisimilar than another; or on what evidential grounds are we to rationally and conjecturally claim that one theory is closer to the truth than another.^{[73] [74]}

According to Zwart (2001: 121), defining the meaning of “approach to the truth” is not only an analytical affair, but is also subject to what Niiniluoto (1987: 265) calls Augustine's objection: “To judge that a son resembles his father presupposes acquaintance with the father; similarly, to judge that a theory resembles the truth pre­supposes that the truth is already known”. This point embodies the important differ­ence between the logical and the epistemological problem: in the former, in order to make a comparison of any two theories with regard to their closeness to the truth, it is assumed that the truth (or, in a sense, the true theory, supposed as existing) is known, whereas in the latter the relative verisimilitude is estimated without know­ing the truth, namely one needs an appropriate notion of estimated verisimilitude in order to compare two theories with regard to their closeness to the unknown truth.

Therefore, any answer to the first kind of question, explaining under what cir­cumstances one theory is closer to the truth than another, presupposes an aprioris- tic complete knowledge of the truth.

For these reasons we are more interested in, and we will try to give a qualitative answer of, the epistemological problem of estimating a possible scientific progress, and not in the logical one of establishing, in a state of omniscience, which of two theories is the more verisimilar. It is probable that in our solution of the epistemo­logical problem a solution of the logical one is implicit. Anyway the important point is that we have to evaluate the truthlikeness of a theory without knowing the truth.

5