Tarski et la suppositio materialis

Abstract

Dans son article de 1944, « The Semantic Conception of Truth and the Foundations of Semantics », Alfred Tarski réfère en propres termes à la notion médiévale de « suppositio materialis ». L’interprétation qu’il en suggère, cependant, est historiquement trompeuse et l’inexactitude historique se double, en l’occurrence, de ce que l’on peut tenir pour une malencontreuse erreur philosophique. Dans « “la neige est blanche” est vraie », Tarski voit l’expression « la neige est blanche » (entre guillemets) comme le nom d’une certaine phrase, alors que les médiévaux y auraient vu plutôt une occurrence de cette phrase elle-même prise dans un usage spécial, la suppositio materialis. L’article montre en quoi les deux approches diffèrent et soutient que la théorie médiévale est philosophiquement préférable, en ce que : 1) elle est descriptivement plus adéquate en ce qui concerne le fonctionnement réel des langues naturelles ; 2) elle est plus appropriée même pour les langages construits qu’elle rend plus fonctionnels et plus intelligibles ; 3) elle repose sur l’identification d’un phénomène important dont la généralité échappe aux sémantiques d’inspiration tarskienne, celui de la dualité de principe entre l’extension d’un terme pris en lui-même et celle qu’il reçoit dans un contexte propositionnel donné.In his 1944 paper “The Semantic Conception of Truth and the Foundations of Semantics”, Alfred Tarski refers in so many words to the medieval idea of “suppositio materialis”. The interpretation he suggests for it, however, is historically misleading, and this historical inaccuracy yields in this case what can be taken to be an unfortunate philosophical mistake. In “ ‘snow is white’ is true ”, Tarski sees the phrase “snow is white” (between quotation marks) as the name of a certain sentence, while the medieval philosophers would have seen it rather as an occurrence of that very sentence, but taken in a special use, the suppositio materialis . The paper shows how these two approaches differ exactly and argues that the medieval theory is philosophically preferable in that (1) it is descriptively more adequate with respect to natural languages, (2) it is more appropriate even for artificial languages, which it renders both more effective and more intelligible, and (3) it rests upon the identification of an important phenomenon, the generality of which is missed by the Tarskian type semantics, namely the duality of principle between the extension of a term in itself and the extension it receives within a given propositional context

John Buridan on the Bearer of Logical Relations

“…It seems that, according to Buridan, once we are in this pretense situation, the determining of the time referred to by the tense marker is automatic, and appeals only to brute facts of a pretense context. 10 This double working of the context, in the determination of what is said and in the evaluation of the truthvalue of what is said, is needed in the understanding of indexicals -it is only natural that it shows up here, as Buridan explains how 'today' should be interpreted in a pretense situation. The underdetermination of the time aimed at by the present tense is a different phenomenon.…”

“…[14] 262. 27 On the role of the intention of the speaker in the determination of material suppositon, in Ockham's theory, see [10]. 28 [5] III, 4, p. 48, 17-23.…”

Richard Brinkley on Supposition

“…The notion of mention implies that a term has been turned into a name for otheroccurrences ofthe same type, whereas medieval material supposition does not imply any nominalization. On that question, see Brands (iggo) and Panaccio (2004).…”