意味論的真理とその病理性について

Abstract

It is known that the concept of truth behaves pathologically. The simple example is the liar sentence, "This sentence is false, " which says reflectively it is false and its vicious circle. It is pointed out that the circularity of truth causes pathological behavior and that there are intuitively paradoxical sentences when the combination of the naive concept of truth and a circular expression such as the liar sentence is present in a language. However, since we indeed use circular expressions in everyday languages, the concept of truth used in our daily communication must behave pathologically. The theory of truth in analytical philosophy is formally founded by Alfred Tarski's semantic theory of truth and formal semantics. Tarski reveals the logical property of truth through the analytical of pathological behavior of truth in order to define an inconsistent truth predicate for scientific investigations. The defect of his theory is that we have to abandon many normative roles of truth in compensation for the inconsistency of truth. Then, every semantic theory of truth following his theory should addresses the two incompatible questions : how do we define the ordinary concept of truth and how do we define the nonpathological truth? In this point, Tarski's formal device to outlaw the circular expression of truth leads us to classify the sentence into two classes, the nonpathological sentence and the pathological sentence. The question will arise from the philosophical points of view what principle we need to classify the sentence. If we can understand the principle, we will give the better account for the pathology of truth and it might help us to understand the concept of truth itself. With reference to these introductory remarks, the first aim of this paper is to survey the three major semantic theories of truth : (1) Tarski's semantic theory of truth, (2) Saul Kripke's fixed point languages, and (3) Anil Gupta's revision theories of truth and consider the principles of classification assumed in each theory. By the examination of Tarski's concept of "semantic/nonsemantic" and Kripke's "grounded/ungrounded", Tarski and Kripke assume the similar principle of classification based on the intuitive distinction between the linguistic facts and the physical facts. Therefore, Kripke's theory of truth suffers the same defect as Tarski's one. This fact leads us to the hypothesis that circularity is the essential property of truth that enables truth to fulfill its normative role, while it caused pathological behavior. On this hypothesis, Gupta advances the new logic and semantics, the revision theories of truth, in order to formalize the circulation of truth. My second aim is to specify the philosophical significance of Gupta's new theory. The rich systems of the revision theories offer the formal foundation to the former solutions of pathological behavior of truth. The semantic interpretation of pathological sentence is governed by the linguistic conventions about the concept of truth. This means that interpretation of pathological sentence is totally independent of the physical facts. It explains the fact that the different interpretations of the truth predicate can be significantly compatible. On the other hand, the core of the semantic interpretation of truth predicate is uniquely determined by the correspondence to the physical facts. If one respects the fundamental intuition that the concept of truth should have correspondence to facts, then he must admit the legitimacy of this interpretation. However, it is not the only legitimate concept of truth. It presents the minimum condition to be fulfilled by all legitimate concept of truth.