L'Architecture des 'Essais' de Montaigne: memoire artificielle et mythologie.

Daniel Martin develops an idiosyncratic but utterly plausible methodology which he terms "mnemocritique," or criticism based on mnemonics. Mnemonics locates textual places invested with meaning which are to be excavated through narrative relationships. Returning to classical rhetoric and its reliance on a form of mental writing, evocative images which function as text, Martin works from a bibliography of predominantly structuralist orientation, focusing on patterns, locations and juxtapositions within the Essais. "The blueprint of the Essais engenders its own text, a tissue tesserated by the mythology which animates it" (34). Martin also relies heavily, although creatively, on the work of Frances Yates on the art of memory. The patterns in the text create meaning through their components and through their placement. In Montaigne, the primary loci of meaning are found through reference to mythological narratives. In this regard, Martin consults in particular the Metamorphoses of Ovid. Martin orders his investigation by constructing a list of gods. Each god is an orienting factor for several essays, and serves as the sign under which those essays are written.

Martin finds that Montaigne shaped his text as a repository of shards of significance, which could be shuffled and resituated to structure a variety of senses; the Essais is a template which encourages modification and variation. He cites Montaigne to this effect: "a sufficient reader often finds in another's writing perfections other than those intended or perceived by the author, and gives to that which he finds a richer meaning and countenance" (169).

Martin reads the structures in the Essais against buildings which he uses to exemplify certain traits also found in Montaigne. Martin describes Jacques Androuet du Cerceau's chateau deforme ovale as "an example of obsessive symmetry" (8). This is a contemporary version of concern for order, detail and agencement which confirms Martin's perception of Montaigne's extreme attentiveness to appropriate ordering in his text. Montaigne's textual construction is equally empirical, rooted in the real world, Martin maintains: "Montaigne's creative process works backwards; he chooses a place about which he wants to speak or in which he wants to make a character speak. He places himself in the interior of this site, from within which he attempts to create a text matched to this location"(112).

Each textual place corresponds to an image. Martin examines the order of Montaigne's chapters to elucidate the places contained within them. His suggests that the images for which one should search are those which stand out because of their oddness; their strangeness ("violence, shock value, bloodiness, obscenity") allows their inscription as carriers of a textual message. "Each textual locus may actually contain several images at once." (46) When this is the case, the composite of images creates its own sort of chronology or narrative.

Montaigne groups his chapters in clusters of five, each cluster being organized spatially "as a painter would deploy his paints on a wall" (47). Martin speculates that the center of each group is its most crucial point, and that significance radiates out from this center. He looks for mirror-effect symmetries rather than meanings which "slide off from" (49), or develop in glissendo, from each other. Each chapter receives its meaning only through juxtaposition with other chapters, indeed through opposition.

Martin's analysis finds most pertinent the links developed in the Essais between Saturn and Mercury. He examines the intertextual dialogue which can be heard among versions of myths in which these two gods feature. Martin treats the Essais as constitutive of a triptych which is subsequently divided into three parts, peopled respectively by "Mercuriens," "Coperniciens," and "Saturniens."

Martin's reading of the Essais, or his wise restructuring, shows that the apparent disorder of the Essais is deceptive. He demonstrates links among chapters, even those that are not in proximity to each other, which evoke a complex series of echoes and responses within the text itself. In so doing, Martin sharply challenges traditional interpretations of the chronology of Montaigne's compositional strategy (e.g. Pierre Villey; cf. 170), maintaining that Montaigne could feasibly have written all of his chapters simultaneously. The "original" edition of the Essais would then no longer be the first printed version, but rather "the mnemonic schema of the chapters, with schematic indications concerning the mythology to be employed" (87). The Essais might then function, Martin suggests, as a reader-friendly, manipulable and plastic "hypertext" (87).

Martin's is a brilliant revisionist reading of the Essais which performs Montaigne's text in a writerly fashion, to use Roland Barthes' description.

Catharine Randall Coats BARNARD COLLEGE