"It will afford profit and pleasure to that numerous class of persons who have no instinctive enjoyment of literature," writes a British reviewer of the French text of S/Z. Instinctive enjoyment of literature! Surely all of Roland Barthes's ten books exist to unmask such an expression, to expose such a myth. It is precisely our "instinctive enjoyment" which is acculturated, determined, in bondage. Only when we know—and it is a knowledge gained by taking pains, by renouncing what Freud calls instinctual gratification—what we are doing when we read, are we free to enjoy what we read. As long as our enjoyment is—or is said to be—instinctive it is not enjoyment, it is terrorism. For literature is like love in La Rochefoucauld: no one would ever have experienced it if he had not first read about it in books. We require an education in literature as in the sentiments in order to discover that what we assumed—with the complicity of our teachers—was nature is in fact culture, that what was given is no more than a way of taking. And we must learn, when we take, the cost of our participation, or else we shall pay much more. We shall pay our capacity to read at all.

Barthes calls his study an essay, and in it a consideration of more than just the tale by Balzac is desirable if we hope to discern what it is that is being tried here. For the work on the text by Balzac, the dissection—into 561 numbered fragments, or lexias, varying in length from one word to several lines—of Sarrasine, is not performed for the sake of identifying the five notorious codes (hermeneutic, semantic, proairetic, cultural, and symbolic), or even for the sake of discriminating the classical text (with its parsimonious plurality of interpretation and its closure of significance) from the modern text which has no such restrictions, no such closure (for the final closure of the modern text is suspension). Rather, the work so joyously performed here is undertaken for the sake of the 93 divagations (I use Mallarmé's term advisedly, for it is with Mallarmé, Barthes has said, that our "modernity" begins) identified by Roman numerals and printed in large type, amounting in each case to a page or two. These divagations, taken together, as they interrupt and are generated by the lexias of the analyzed text, constitute the most sustained yet pulverized meditation on reading I know in all of Western critical literature. They afford—though Barthes can afford them only because of the scrupulous density of his attention, his presence of mind where one is used to little more than pasturage—a convinced, euphoric, even a militant critique of what it is we do when we read. (pp. ix-x)

"What do you read now?" the hungry interviewer asked the famous writer, a woman of commercial success in the theater whose autobiography has defined a character of considerable literary sophistication. And the famous writer answered:

I don't read novels any more, I'm sorry to say. A writer should read novels. When I do, I go back to the ones I've read before. Dickens. Balzac … I find now when I go to get a book off the shelf, I pick something I've read before, as if I didn't dare try anything new.

Aside from the underlining fact that it is a writer speaking, this is a familiar experience, this preference for what Barthes calls the readerly over what he calls the writerly…. It is a familiar experience because only what is authentically writerly can become readerly. If we were to set out to write a readerly text, we should be no more than hacks in bad faith; yet, as readers, how hard it is to face the open text, the plurality of signification, the suspension of meaning. It explains that hesitation at the bookshelf, the hand falling on the Balzac story, the known quantity. Known … How often we need to be assured of what we know in the old ways of knowing—how seldom we can afford to venture beyond the pale into that chromatic fantasy where, as Rilke said (in 1908!), "begins the revision of categories, where something past comes again, as though out of the future; something formerly accomplished as something to be completed." (A perfect description, by the way, of the book in hand.) Why we read in this repressed and repressive way; what it is, in the very nature of reading, which fences us in, which closes us off, it is Barthes's genius to explore, not merely to deplore. His researches into the structure of narrative have granted him a conviction (or a reprieve), a conviction that all telling modifies what is being told, so that what the linguists call the message is a parameter of its performance. Indeed, his conviction of reading is that what is told is always the telling. And this he does not arraign, he celebrates.

So exact are Barthes's divagations, so exacting are their discoveries about the nature of reading, that we may now and again be dismayed—if we are in the main readers of the readerly—by the terms he has come to (he usually assumes Greek has a word for it) in which they must be rendered. For Barthes's text is writerly—at least his divagations are. This criticism is literature. It makes upon us strenuous demands, exactions. And because of them, precisely, we too are released, reprieved; we are free to read both the readerly (and can we ever again read Balzac in all innocence? can we ever want to?) and the writerly, en connaissance de cause, knowing the reason why. Essentially an erotic meditation, then, because it concerns what is inexpressible (which is the essence of eros), Barthes's essay is the most useful, the most intimate, and the most suggestive book I have ever read about why I have ever read a book. It is, by the way, useful, intimate, and suggestive about Balzac's tale Sarrasine, which the reader of the readerly will find reassembled at the end of this writerly book, en appendice, as the French say. (pp. x-xii)