376 EIGHTEENTH-CENTURY FICTION 3:4 Les essais de Kibédi Varga, Gerald Prince, Michèle Weil, et Ian Lancashire constituent des réflexions théoriques sur le concept de topos. La convergence de leurs conclusions offre une vision plus rigoureuse de la nature et du fonctionnement des topoï romanesques. Le lecteur y trouvera un classement des différents types de lieux, un aperçu des rapports entre topoï et dénarré, celui-ci se présentant comme "centre d'accueil privilégié" par sa fonction interprétative, une description du topos comme "configuration narrative récurrente," permettant son repérage et une mise au point des problèmes que pose encore la recherche informatisée sur le topos. Maria Joâo Brilhante Université de Lisbonne Jean Goldzink. Charles-Louis de Montesquieu: Lettres persanes. Paris: Presses universitaires de France (Etudes littéraires, N0 23), 1989. 123pp. "Nos ancêtres étaient gens un peu rudes," Jean Goldzink writes (p. 91), referring to Daniel Mornet and other older critics who, neglecting the fictional element in the Lettres persanes, read the work as an account of Montesquieu's opinions (summed up as "relativism ") on a number of philosophical topics. However, since about 1960, "grâce aux progrès de la technique et de la coopération internationale, nous ne sommes plus aussi naïfs" (p. 91); in other words, Montesquieu's early masterpiece is now accepted as a work of prose fiction—indeed, some modern critics see it as a fully fledged novel with a coherent plot (the harem story) and psychologically complex characters. Goldzink, though loudly congratulating himself on being a "modern," makes no such extravagant claim; he speaks of "la double nature du livre: mixte de philosophie et de récit" (p. 31); his aim is to do justice to the hybrid nature of the work. Goldzink argues that récit (or story) and ideas are not separate or separable compartments : the story is essentially the Persians' voyage of discovery and its consequences, which include not only the revolt in the harem but also the travellers' reactions to the Western way of life. This is a story which generates ideas and, conversely, the ideas are dramatized in the story, for the experiences of Usbek and Rica are related in a series of dated letters which must be read in context—in the context of time and place, and also in the context of destinateur and destinataire. This is the heart of the matter: Montesquieu is not speaking in his own voice, but inventing a multi-voiced fictional correspondence in which he orchestrates a rich variety of epistolary strategies where the status of any particular énoncé is highly ambiguous. Readers, confronted with "les rouages de l'ingénieuse machine textuelle imaginée par Montesquieu" (p. 94), inevitably have difficulty in finding their bearings in a work which dissolves into a cacophony of conflicting voices. There is a point in his book where Goldzink, having expelled the author and shattered the text into a multiplicity of subjective views, seems to accept that we should simply revel in "le plaisir du texte": "Pourquoi ne pas lire les Lettres persanes comme une juxtaposition de discours, philosophiques, religieux, satiriques, qui se succèdent, se croisent, se détruisent, sans qu'il y ait à chercher une voix d'auteur, des valeurs assum ées, parce que le sens du livre se confond avec le tressage même des discours?" (p. 42). Here we seem to be at the opposite pole from the older academic criticism; however , in the next sentence Goldzink makes the surprise announcement that beneath the motley surface of the Persians' correspondence, it is possible, after all, to discover the REVIEWS 377 voice of the author, whose philosophy is that, at the deepest level of human experience , there is a rational moral order. To demonstrate this, Goldzink follows a procedure not easily distinguishable from Mornet's, which is to assume that certain "enlightened" ideas expressed by the Persians are endorsed by Montesquieu. Thus he refers to letter 83, where Usbek offers a rational proof of the existence of God and Natural Law; but the reader will surely feel that this tells us nothing about Montesquieu's...