Introduction

Does science progress, and can its progress be estimated? Is our knowledge cumu­lative, so that our theories progress towards some paramount cognitive goal, such as some sort of truth? Are some of our ideas about the world more rational, more scientific than others? Are our beliefs on nature, on its mechanisms, really worthy of credence?

Questions like these about the nature of scientific knowledge—the kind of knowledge that is shaped into scientific theories and consists of the ascertainment, understanding and explanation of the phenomena of the world—are much less

V.

Department of Basic Sciences and Foundations,

University of Urbino ‘Carlo Bo’, Urbino, Italy

e-mail: vincenzo.fano@uniurb.it

G. Macchia

e-mail: lucbian@hotmail.com

© Springer International Publishing Switzerland 2015 65

M. Alai et al. (eds.), Science Between Truth and Ethical Responsibility,

DOI 10.1007/978-3-319-16369-7_5

naive, and call for answers much less plain than might appear. And, although the progressiveness of science, its rationality, and its cumulative path, seem in a sense obvious facts, probably taken for granted not only by most laymen, but also by the majority of scientists, actually, these questions still remain unanswered, tricky, and therefore of broad and current interest among philosophers of science.¹

In this respect, it suffices to loosely consider the progress of knowledge in dif­ferent ages of our history: it is reasonable to think that today the international sci­entific community has more knowledge than a hundred years ago, but it is also unquestionable that the scientific (in a broad sense) community in the second cen­tury before Christ possessed more knowledge than in the seventh after Christ. So scientific progress is not a necessary path of history,^{[63] [64] the time arrow of knowl­edge—so to speak—is not at all unidirectional, so also a scientific regress in knowledge is possible.}

Consequently, concepts like “theory change”, “cumulative growth”, “truth”, “true theory”, “verisimilitude”, “truthlikeness”, “belief revision”, all belonging to that field of research named theory of scientific progress and all trying to capture the elusiveness of our knowledge dynamics, have proved to be particularly slip­pery and controversial.

In this paper, we will be dealing with one of these elements: the problem of verisi­militude (usually also called truthlikeness or approximate truth).^[65] The notion of the verisimilitude of a theory attempts to explicate how close a theory is to the truth. In other, more intuitive, terms, this notion can be expressed by saying that a theory is more verisimilar than another when “the first says more things about the domain under investigation and more of all things said are true” (Kuipers 2014: 64). Therefore, the problem of verisimilitude may be summarized with a question like this: What can there be about a theory that makes it closer to the truth than another one?

On the other hand, we know that every scientific theory, however established and cherished, is presumably false, that is, with at least one false consequence. Any theory, indeed, is an idealization and an oversimplification of the real world, and, sooner or later, it will probably be included in some other more general the­ory, insofar as a perfect match between theories and the world is actually, and for many reasons, impossible (just think, for instance, about the unobservable and spatiotemporally remote aspects of the universe). However, some (false) theo­ries objectively seem to be better than others. Let us consider, as a very simple example, the theories of motion: starting with Plato and Aristotle, they then devel­oped through Philoponus, Buridan, Oresme, Galileo and Descartes, until arriving at Newton and finally Einstein, so giving the appearance of a real improvement: Einstein’s theory, though presumably false, is more successful than its predeces­sors essentially because it is closer to the truth, that is, nearer to the unknown true theory about motion.

It should not be overlooked that the concept of truth is, as Oddie (2014) rightly points out, a rather coarse-grained property of propositions, so a more fine-grained ordering, such as the degree of closeness to the truth, is naturally compelling. After all, such a notion is really basic, given also that its importance follows from two modest and minimal realist assumptions, widely accepted: “The truth doc­trine (that the aim of an inquiry, as an inquiry, is the truth of some matter) and the progress doctrine (that one false theory may realise this aim better than another). Together these yield the conclusion that a false theory may be more truthlike, or closer to the truth, than another” (Oddie 1986: ix). Furthermore, and still more significantly, Oddie adds: “Truthlikeness is not only a requirement of a particular philosophical outlook, it is as deeply embedded in common sense as the concept of truth. Everyone seems to be capable of grading various propositions, in differ­ent (hypothetical) situations, according to their closeness to the truth in those situ­ations” (ibid.).

Notwithstanding, it is important to remark that there are research programmes, such as the belief revision programme, that do not consider truth and falsity as necessary means for an adequate understanding of the methodological rules that guide the change of beliefs and consequently of the nature of science,^[66] or also pro­grammes which are against the idea that science progresses as it gains more verisi- militude.^[67] Indeed, outstanding philosophers with pragmatist leanings, such as Thomas Kuhn and Larry Laudan, maintain the idea that science is not a truth-seek­ing enterprise but rather a problem-solving activity.

Our paper proceeds as follows. In Sect. 2 we very briefly hint at Popper’s his­torical proposal on verisimilitude that gave birth to this field of research. In Sect. 3 we introduce Agazzi’s approach, which is our starting philosophical basis. After dwelling, in Sect. 4, upon the distinction between logical and epistemological approaches to the verisimilitude problem, specifying our preference for the latter, in Sect. 5 we take into account the traditional syntactic view of theories. Its weak points lead us to adopt the semantic view of theories in Sect. 6, where we will enter into the details of the notions and of the constitutive elements of this view, also developing our epistemological proposal about the comparative evaluation of theories and cognitive situations. Finally, Sect. 7 offers our conclusions.

2