Martine Pécharman

Pascal’s wager was sometimes viewed in the eighteenth-century as an argument conquering its whole demonstrative force not in the Pensées but in a passage of Locke’s Essay Concerning the Human Understanding (II, XXI, § 70) dealing with the preference to be given to a virtuous life when considering the possibility of another eternal life. In this paper, I intend to show that this interpretation is ill-founded. The argument of the wager highlights the discrepancy between the requirements of alethic reason on the one side and of practical reason on the other side: the wager is to be seen not as a proposition, but as an act. When emphasizing the conditions on which this act can conform to rationality, the argument used of in the Pensées puts the comparison between a unity (my life ante mortem) and a multiplicity (other possible lives post mortem) on the front of the stage, in combination with a reckoning of the chances of winning or losing the bet on God’s existence. The reduction ad absurdum of the act of betting against God’s existence is deduced from a double infinitization, not only quantitative (an infinity of lives), but also qualitative (a reiterating infinity of happiness) of the second member in the comparison. Now, the case is different in Locke’s argumentation. Its reading is vitiated if we deem that the § 70 in the Essay II, XXI consists in a self-sufficient demonstration, understandable by itself. Through a close examination of the chapter XXI in its entire, it appears on the contrary that the thesis of an alleged improvement or even of a mere equivalent of Pascal’s wager in the Essay is defeated by the necessity to put back Locke’s own argument in its true place, actually as the result of a complex enquiry on the principle of the determination of the will.

The addition to the fifth edition (1683) of La Logique ou L’Art de penser (1662), first a chapter dedicated to the general definition of a sign, then five chapters specifically relating to an analysis of linguistic signs, has sometimes been interpreted as a belated appearance of the “problem of language” in the Logique. Because most of these supplementary chapters had been patched together from passages copied from the Perpétuité de la foi (1669-1674), this has led commentators to attribute the later foregrounding of the “problem of language” in the Logique to a semiotic analysis that arises in Port Royal’s anti-Calvinist polemic over the meaning of “Hoc est corpus meum”.Against this reading, in the first part of my essay, I try to show that from the first edition of L’Art de penser, the analysis of language had been inseparable from the analysis of the operations of the mind. For Port Royal, the central function of the Logique involves resolving all instances of linguistic confusion or equivocation which disturb or prevent grasping the speaker’s meaning. The aim of the logical analysis of L’Art de penser is to train the judgment to understand linguistic expressions that involve gaps and deviations with respect to a model of term by term correspondence between ideas and words. In the second part of the essay, I oppose the thesis that the theorization of sign and signification in the chapters added to the 1683 edition of the Logique depends on the theological dimension of the anti-Calvinist discussion. I show that the critique of the Calvinist conception of the Eucharist as the sign of the absent body of Christ in fact itself constitutes in the Perpétuité de la foi an application of the analysis of language proposed in 1662 in the Logique. That which the enlarged edition of 1683 incorporates into L’Art de penser from the Perpétuité de la foi thus constitutes the part of the anti-Calvinist controversy that is itself logical, not theological, the extension of the reflections of 1662 on the difficulties of relating that which is said to that which is meant. A third part of the essay is devoted to the construction of the general definition of a sign in the 1683 edition of the Logique. On the one hand, I reconstitute the way in which Port Royal proceeds to a reconstruction of the definition of Augustine’s De doctrina christiana. On the other hand, I show that this general definition grounds an identification of the relation of linguistic signification with a relation of representation between words and ideas, which reduplicates in 1683 the relation of representation of ideas to things established in 1662.