The Referential Nature of Truth

The difference and interrelation between sense and reference was recognized already in ancient philosophy in the efforts of defining truth, and was explicitly pointed out in the Aristotelian distinction between semantic discourse (i.e.

the discourse that is endowed simply with meaning), and apophantic discourse (i.e. the discourse that affirms or denies something) because in this second case one has to consider “about what” the affirmation or negation is stated, and on this ground the discourse turns out to be true or false. To put it differently, it is not granted that whatever meaningful declarative discourse is true: in addition it must “say of” something what this something “really is”. This is the familiar commonsensical notion of the truth of a statement that has also been accepted in philosophy until the twentieth century (and which underlies the different criteria for truth proposed by the different “theories of truth”). In the case of the sciences, the ‘crisis' occurred especially at the beginning of the twentieth century (of which we have spoken in Sect. 3 of the present paper) had led many epistemologists to consider scientific theories essentially as formal systems whose global context provides at the same time the sense and the content of their statements, provided that they be internally consistent (i.e. free from contradiction). This was a kind of “syntactic conception of truth” or “coherence theory of truth”, whose weakness became patent especially after Godel's result concerning the impossibility that any formal system (satisfying certain minimal conditions) can prove its own consistency. This produced the challenge to reintroduce the traditional proper notion of truth also for the formalized languages, and this result was offered in the famous paper by Tarski (1933)8 who explicitly wanted to qualify his doctrine as a “semantic conception of truth”.^[15] ^{^[16] If one considers the core of Tarski's very complex and elaborated construction one recognizes two things: (a) that the “interpretation” of the formal language does not consist in connecting its signs with concepts or giving them a sense, but in connecting them directly with unqualified elements of a given set (i.e. with referents), and this was the reason for calling his conception “semantic”; (b) the necessary and sufficient condition for declaring the truth of a proposition is that the state of affairs described by the proposition actually holds, but no criteria are offered for checking whether or not such a condition is fulfilled.
These can perhaps be considered weak points or at least limitations of the Tarskian definition of truth. Two fundamental acquisitions, however, must be recognized: (a) to have recovered the referential nature of truth, that is, the inadequacy of considering it as a purely internal property of a linguistic construction; (b) the ‘bilateral' condition for the truth of a declarative sentence, that is: if a state of affairs (or a fact) obtains, then the proposition describing it is true, and if a proposition is true, then the state of affairs (the fact) it describes must obtain. One can discuss whether or not this Tarskian conception can be qualified as a “correspondence theory” of truth; this discussion is rather idle because it is not univocally understood what such a correspondence should be. This is why we prefer to call it “referential”. Not just for the sake of clarity, but also because we have already indicated how one can trace the referents of linguistic elements, since our operations constitute at the same time “criteria of reference” and “criteria of truth”, as we have seen. 8}

<< | >>

↑

Source: Agazzi E. (ed.). Varieties of Scientific Realism: Objectivity and Truth in Science. Springer,2017. — 411 pp.. 2017

The Referential Nature of Truth

More on the topic The Referential Nature of Truth: