...Commercial realism has cornered the market, has become the most powerful brand in fiction. We must expect that his brand will be economically reproduced, over and over again. That is why the complaint that realism is no more than a grammar or set of rules that obscures life is generally a better description of le Carre or P. D. James than it is of Flaubert or George Eliot or Isherwood: when a style decomposes, flattens itself down into a genre, then indeed it does become a set of mannerisms and often pretty lifeless techniques. The efficiency of the thriller genre takes just what it needs from the much less efficient Flaubert or Isherwood, and throws away what made those writers truly alive. And of course, the most economically privileged genre of this kind of largely lifeless "realism" is commerical cinema, through which most people nowadays receive their idea of what constitutes a "realistic" narrative.
Decomposition like this happens to any long-lived or successful style, surely; so the writer's--or the critic's, or reader's--task is then to search for the irreducible, the superfluous, the margin of gratuity, the element in a style which cannot be easily reproduced and reduced.
But rather than do that, Barthes and Moody and Giles and William Gass and many other opponents of fictive convention confuse two quite different complaints. Here is Barthes in 1966: "The function of narrative is not to 'represent,' it is to constitute a spectacle still very enigmatic for us but in any case not of a mimetic order . . . 'What takes place' in the narrative is, from the referential (reality) point of view literally nothing; 'what happens' is language alone, the adventure of language, the unceasing celebration of its coming" Now, to charge fiction with conventionality is one thing; to move from this charge to the very skeptical conclusion that fictive convention can therefore never convey anything real, that narrative represents "literally nothing," is incoherent. First, all fiction is conventional in one way or another, and if you reject a certain kind of realism for being conventional, you will also have to reject for the same reason surrealism, science fiction, self-reflexive postmodernism, novels with four different endings, and so on. Convention is everywhere, and triumphs like old age: once you have reached a certain seniority, you either die of it, or with it. One of the nice comedies of Cyril Connolly's essay is that by blacklisting every conceivable convention he effectively bans the writing of any fiction at all--"anybody over six feet, or with any distinction whatever." Second, just because artifice and convention are involved in a literary style does not mean that realism (or any other narrative style) is so artificial and conventional that it is incapable of referring to reality. Narrative can be conventional without being a purely arbitrary, nonreferential technique like the form of a sonnet or the sentence with which Snoopy always begins his stories ("It was a dark and stormy night . . .")
Paul Valery was hostilein a Barthes-like way to the claims of fictional narrative, and his example of an entirely arbitrary fictional premise was a sentence like: "The Marquise went out at five o'clock." Valery felt, as William Gass did when discussing James's Mr. Cashmore, that this sentence is exchangeable with an infinite number of other possible sentences, and that this kind of provisionality robs narrative fiction of its necessity and its claim to probability. But as soon as I place a second sentence on the page--"That letter, received in the morning, had irritated the Marquise, and she was going to do something about it," say--the first sentence no longer looks quite as arbitrary or peremptory or merely formal. A system of relations and affiliations is beginning to quicken. And as Julien Gracq points out, "Marquise" and "five o'clock" are not arbitrary at all, but full of limit and suggestion: a marquise is not an ordinary, interchangeable citizen, and five o'clock is still late afternoon while six is drinks time. So what is the Marquise going out for?
The point to make about convention is not that it is untruthful per se, but that it has a way of becoming, by repetition, steadily more and more conventional. Love becomes routine (and indeed Barthes once claimed that "I love you" is the most cliched thing anyone can say), but falling in love is not nullified by this fact. Metaphors become dead through overuse, but it would be insane to charge metaphor itself with deadness. When the first caveman, shivering, said that he was as cold as ice, his interlocutor probably exclaimed: "That is pure genius!" (And after all, ice is cold.) Likewise, if someone were now to paint in the style of Rembrandt, he would be a third-rate copyist, not an original genius. These are the simplest arguments, and one should not have to make them, were it not for a persistent tendency among those hostile to verisimilitude to confuse convention with an inability to refer to anything truthful at all
Brigid Lowe argues that the question of fiction's referentiality--does fiction make true statements about the world?--is the wrong one, because fiction does not ask us to believe things (in a philosophical sense) but to imagine them (in an artistic sense): "Imagining the heat of the sun on your back is about as different an activity as can be from believing that tomorrow it will be sunny. One experience is all but sensual, the other wholly abstract. When we tell a story, although we may hope to teach a lesson, our primary objective is to produce an imaginative experience." She proposes that we restore the Greek hetorical term "hypotyposis," which means to put something before our eyes, to bring it alive for us. (Somehow I don't think that "hypotyposis" will displace "realism" as the preferred term any time soon.)
If we reexamine Aristotle's original formulation of mimesis, in the Poetics, we find that his definition is not about reference. History shows us, says Aristotle, "what Alcibaides did"; poetry--i.e., fictional narrative--shows us "the kind of thing that would happen" to Alcibaides. Hypothetical plausibility--probability--is the important and neglected idea here: probability involves the defense of the credible imagination against the incredible. This is surely why Aristotle writes that a convincing impossibility in mimesis is always preferable to an unconvincing possibility. The burden is instantly placed not on simple verisimilitude or reference (since Aristotle concedes that an artist may represent something that is physically impossible), but on mimetic persuasion: it is the artist's task to convince us that this could have happened. Internal consistency and plausibility then become more important than referential rectitude. And this task will of course involve much fictive artifice and not mere reportage.
So let us replace the always problematic word "realism" with the much more problematic word "truth" . . . Once we throw the term "realism" overboard, we can account for the ways in which, say, Kafka's Metamorphosis and Hamsun's Hunger and Beckett's Endgame are not representations of likely or typical human activity but are nevertheless harrowingly truthful texts. This, we say to ourselves, is what it owuld feel like to be an outcast from one's family, like an insect (Kafka), or a young madman (Hamsun), or an aged parent kept in a bin and fed pap (Beckett). There is still nothing as terrifying in contemporary fiction, not even in the blood-bin of Cormac McCarthy or the sadistic eros of Dennis Cooper, as the moment when Knut Hamsun's narrator in Hunger, a starving intellectual, puts his finger in his mouth and starts eating himself. None of us, I hope, has done this, or will ever want to. But Hamsun has made us share it, has made us feel it. Dr. Johnson, in his "Preface to Shakespeare," reminds us, "limitations produce pain or pleasure, not because they are mistaken for realities, but because they bring realities to mind."
Convention itself, like metaphor itself, is not dead; but it is always dying. So the artist is always trying to outwit it. But in outwitting it, the artist is always establishing another dying convention. It is this paradox that explains the further, well-known literary-historical paradox, namely that poets and novelists repeatedly attack one kind of realism only to argue for their own kind of realism.
James Wood - How Fiction Works