A decade ago, Susan Sontag suggested that rather than an esthetics we need an erotics of art. Roland Barthes has gone some way toward providing this in a slim volume called "The Pleasure of the Text" ["Le plaisir du texte"], a liberated and self-indulgent meditation on the solitary vice of reading. Here, and in the book that immediately preceded it, "S/Z," Barthes demonstrates a renewed suppleness that takes him beyond the limitations of much recent French work of a "structuralist" persuasion, including his own.

Taken together, "The Pleasure of the Text," and "S/Z" force us to notice how much of the most interesting thought today is being carried forward in what we used to call "literary criticism," and how important Barthes's own contribution to redefinition of the field has been….

"The Pleasure of the Text" is consciously an assertive book, Nietzschean in its manner, aiming at effect rather than persuasion. It is seductive and liberating, but also a bit overripe, in its discursive value somewhat limited. In "S/Z," on the other hand, we find the hedonist locked in a close and exacting dialectic with the ascetic, and the result is Barthes's best, most enlivening book to date, a work both of systematic analytical rigor and of consciously sensuous pleasure, which claims our attention as a rarely penetrating reflection on how and why we read. Barthes takes a tale by Balzac, "Sarrasine," divides it up into 561 "lexias" or short consecutive units of reading, then traces through each unit, minutely, locally, the interweaving of the five "codes" or "voices" that constitute the text's system of meaning. Then, in the interstices of this painstaking analysis, he unfolds 93 excurses …, episodes of interpretation and theory concerning the emergent patterns of sense, and the process of sense-making. On the one hand, then, discipline, rigor, dissection; on the other, imaginative play and speculative flight: the Thanatos and Eros of criticism in illuminating contest. This gives a book that advances decisively beyond the rather inert mechanism of most structural analysis of narrative, toward a description of the force-field of meaning and the polyphonal richness of reading.

In this book his concerns have converged on the central and difficult problem that more and more appears basic to any critical discourse: the theory of reading. What in fact goes on in the encounter of a reader, himself already a structure of cultural codes and conventions, with the structures and coded messages of the text?

If there have been all along in Barthes two warring selves, the theoretical ascetic (represented by the systematic semiological studies) and the hedonist, indulging himself with reading, it's clear that the hedonist wins hands down in "The Pleasure of the Text." Here, what interests him in the meeting of reader and text is specifically the "dialectics of desire" by which animation of the text depends on the reader's desire, and this on the text's solicitation of the reader. Writing is the amorous science of language ("its Kama Sutra") and reading resembles the floating attention of psychoanalysis, lying in wait for the signs of arousal in and from the text. From pleasure, these can in the most radical instances become orgasmic "bliss": the moments in which the text is most subversive, playful, generative of new possibilities, least wedded to ideology or representative function. Barthes values those texts and textual moments that most appear scandalous, when the reader can accede to the pure play of language in different points in the narrative, and for understanding their role as armature of the story. This code works in close conjunction with the "hermeneutic," which involves the questions or enigmas posed by the story—their suspension, delay, unveiling, resolution. Together, the application of these two codes leads beyond a static conception of form to a dynamic conception of the force of the text as read—and as desiring to be read. The other three codes are perhaps more ordinary pieces of the critical workshop; what's valuable is the way Barthes has grafted them to the structural logic of the narrative, to suggest for instance the place of "character" within the "economy" of textual connotations.

One may resist the claim that the codes cover the field of meaning. But on the whole they do work. They promote an acute and playful reading of "Sarrasine" and, more important, they provoke in the reader a disquieting awareness of the act he is engaged in and the desire that moves him. He finds himself playing with the codes, playing against them sometimes, playing from them. The codes induce a search not for structure, but for structuring: not for final organization of the work but for the fullest productivity of its different registers. Barthes finds an analogy in the musical score: the codes, through the successive "measures" (the lexias), produce meaning in different ranges of the textual instrument, sustain melodies and harmonies, create tonal patterns and cadenzas. Where the practice of literary criticism is concerned, this suggests that Barthes has found a way to answer the call (of Northrop Frye, among others) for a general structure that would permit different critical languages (thematic, rhetorical, psychoanalytic, etc.) to complement one another, in a radical pluralism.

"The plural" of the text is indeed Barthes's insistent subject in both of these books. The plural means not simply "ambiguity" or multiple meanings, but the impossibility of hierarchizing the different voices of the text, the need to submit to their interdependence and simultaneity. Reading should never become the search for some ultimate signified, but rather a joyous and animating multiplication of the signifiers, of words and symbols themselves. For Barthes, there is a distinction to be made between the "parsimonious plural" of a "classic" text (such as Balzac's) and the total plural of such as the "new novelists." The distinction is surely overdrawn—and to most readers the controlled plural of something like "Sarrasine" will be richer than the exercises of such Barthesian admirations of Philippe Sollers. Barthes's insistence upon the overwhelming value of the latest Parisian experimental writing is an annoying provincialism that sometimes leads him into a perverse blindness to the continuities of the literary problems and ambitions that most interest him….

We can be irritated by some of Barthes's assertiveness, his lack of balance as a critic, while recognizing that at his best he provokes with a rare pertinence—as in [the] description (from "The Pleasure of the Text") of the "circular memory" that makes any textual fragment remind us of others, of the web of the written…. The claim made here for the omnipresence of textuality, and hence for the importance of deciphering all the codes of the world, both in suspicion and with pleasure, may appear extravagant, yet also true. The concern with sign-systems is finally a concern with deciphering our culture and ourselves.