Here are two recent statements about literary realism, statements so typical of their age that a realist novelist would have been proud to have imagined them into life. The first is by Rick Moody, reviewing JM Coetzee's novel Elizabeth Costello in Bookforum: "It's quaint to say so, but the realistic novel still needs a kick in the ass. The genre, with its epiphanies, its rising action, its predictable movement, its conventional humanisms, can still entertain and move us on occasion, but for me it's politically and philosophically dubious and often dull." The second is by Patrick Giles, contributing to a long, raucous discussion about fiction, realism, and fictional credibility on a literary blog called The Elegant Variation: "And the notion that this [the realistic novel] is the supreme genre of the lit tradition is so laughable that I ain't even gonna indulge myself."
A style unites the two statements, a down-home relaxation of diction ("kick in the ass," "ain't even gonna"), which itself informs us about the writers' attitudes towards realism's own style: it is thought to be stuffy, correct, unprogressive, and the only way even to discuss it—"so laughable"—is to mock it with its stylistic opposite, the vernacular. Realism, it seems, is so conventional it is almost embarrassing ("quaint") to be caught discussing it at all. A curious anxiety, when one considers that the era Giles dismisses stretches from, at least, Balzac to Forster, from 1830 to 1910. It is a little like dismissing as beneath comedy the idea that English poetry reached any kind of pinnacle from Shakespeare to Milton, or that music did the same from Beethoven to Mahler.
The major struggle in American fiction today is over the question of realism. Anywhere fiction is discussed with partisan heat, a faultline emerges, with "realists" and traditionalists on one side and postmodernists and experimentalists on the other. No comparable struggle exists in British fiction because experimental fiction has never been substantial enough to mount a decent campaign against the dominant discourse. But the 1960s avant-garde in America was full of talent and vigour. In addition to writers like John Barth and Gilbert Sorrentino, who never really reached popular audiences, many of the avant-gardists of that period became mainstream, notably Thomas Pynchon and the delightful story-writer Donald Barthelme and William Gass, and the unclassifiable Kurt Vonnegut. The heirs of this era of experiment might include Don DeLillo, Rick Moody, David Foster Wallace, Paul Auster, Lydia Davis and Ben Marcus, all very different from each other and of different ages, but all committed in one way or another to going beyond realism. A testament to the success of avant-gardism in America was offered, in 1986, by Philip Roth's The Counterlife, which took just what it needed from postmodern narrative games in order to make a fundamentally metaphysical argument about the different ways of living, and narrating, a life. In a younger generation, Jonathan Franzen's writings about whether he is a highbrow artist or a popular entertainer, and his tortured negotiations with the legacies of DeLillo and William Gaddis, are difficult to imagine without the challenge of American experimentalism in the 1960s and 1970s.
There are at least two other large reasons for the vitality of this struggle in American fiction. One has to do with the anti-intellectualism of a good deal of American writing, and the other with the perceived traditionalism of creative writing programmes, always suspected of exerting a grey monopoly over American writing. The two complaints may be linked. The "hardboiled-dom" that Saul Bellow named as an obstacle to mental life on the first page of his first novel, Dangling Man, in 1944, has tended to combine male reticence with an artisanal commitment to "craft"—the latter being, necessarily so, generally a matter of mastering a set of realist techniques. The separation of creative writing programmes and English departments is generally profound in most colleges; one does technique, the other does ideas. It's dismaying to find the richly talented novelist and short story writer Richard Yates, in Blake Bailey's biography, installed at Iowa in the 1970s, almost dead from booze and cigarettes, and dismissing not only Bellow and Roth but even poetry for being too airy-fairy. Yates's marvellous stories have, accordingly, been promoted by male traditionalists like Richard Ford, André Dubus and Richard Russo, though they rightly belong to everyone.
It's difficult to know which side is the more myopic; what is certain is that both camps have a niggardly and often confusedly literal conception of realism. Moody's three sentences efficiently compact the reigning assumptions. Realism is assumed to be a "genre" (rather than, say, a central impulse in fiction); it is taken to be mere dead convention, and to be related to a certain kind of traditional plot, with predictable beginnings and endings; it deals in characters, but softly and piously ("conventional humanisms"); it assumes that the world can be described with a naively stable link between word and referent ("philosophically dubious"); and all this will tend toward a conservative or even oppressive politics ("politically and philosophically dubious"). This might plausibly describe a contemporary novel by Anne Tyler or Kent Haruf, but it is almost an exact inversion of the 19th-century realist novel, which was often politically and philosophically radical. Often, and most notably in Flaubert, it overwhelmed the world with words, with elaborations of style, even as it claimed exactly to match word with referent; and often it dealt savagely and pessimistically with its fictional characters.
It is impossible to separate postmodernism's scepticism about realism from academic literary theory's scepticism, and in the case of the nouveau roman the two belong together like bubbles and champagne, in one rare intoxicating blend of the academic and the literary. Literary theory has been recurringly obsessed with traditional realism, determined to unmask its naivety, proclaim its enthralment to bourgeois convention and gloat over its final impossibility. Thus the paradox that the most sensitive modern analyst of realism, Roland Barthes, the critic who most acutely laid bare realism's grammar with something like a lover's devotion, was almost derangedly hostile to its ambitions, an implacable denier of its effects.
Just as Barthes's devoted hostility to realism seems contradictory, so more broadly there is the paradox that theory has on the one hand unquestionably complicated our sense of realism's procedures but on the other has rarely credited that now enriched sense of complication to the realist writers themselves. Rather, the new complications are credited merely to the texts, whose problematic richness is separated from authorial intention and ascribed to a hapless textual unconscious—or even better, ascribed to the theorist.
Theory, with its suspicious hygiene, has made Balzac's and Zola's novels much more complex than they seemed to earlier generations of readers. No one any longer takes at face value those writers' proclamations that their books will "show everything," will strip reality of its encrusted obstructions and give us reality as it really is, through a pane of glass. We are all alert, now, to the distortion that all style, even the "simplest," exerts on its subject; we have been trained to see descriptive realism, from Balzac to Isherwood, from Flaubert to Graham Greene, as a set of conventions, a breviary of formulas that obscure and constrain as much as they reveal and liberate. For us, the pane of glass is always a bit cracked and is smeared with the writer's fingerprints.
In his great book S/Z, Barthes broke open "Sarrassine," a story by Balzac, and brilliantly demonstrated that its narration rests on a family of conventions, a system of codes, a "grammar" of realism. This grammar is a way, Barthes argues, of structuring not reality but storytelling itself. There is no "realistic" way to narrate the world. The 19th-century author's naive delusion that a word has a necessary and transparent link to its referent has been nullified. We move merely among different, competing genres of fiction-making, of which realism is merely the most confused, and perhaps the most obtuse because least self-conscious about its own procedures. In Barthes's hands, Balzac seems as artificial, as conventional as, say, John le Carré's "realist" storytelling seems to us today: a tissue of signs.
How does the "realist" respond to this? The most obvious objection to the postmodern superiority about realism's "conventionality" is: why should convention make reference to reality impossible? All modes of reference are conventional, including language itself, but we don't assume that such conventionality renders the world indescribable. There is no writing without convention; the struggle is always to be alive to the moment when a literary convention becomes dead. Hemingway told us something new about how people talk; but today any television writer knows how to "do" Hemingway, and a once-revealing convention is now an obscuring one.
But if people in the world really do talk as they do in Hemingway, why would such a convention ever age and get stale? Surely it would be eternally alive? Yes, but people do not talk exactly as they do in Hemingway. Hemingway took a tendency in speech and made a plausible invention of it. He took a match and he showed that you could cook with it; he taught us that a flame can look like this. The convention is dead precisely when it has become so widespread, among writers, dramatists, television writers, that it seems to be saying: people always sound like this. They do not. Once partially true, it has become widely untrue.
The anti-realist might say that this is simply grist to his mill, since I'm suggesting that fiction is just a matter of more or less untruthful conventions. But it would be more honest to claim that fiction is a matter of more-or-less truthful conventions. In fact, anti-realist theory has managed have it both ways: on the one hand realism is excoriated for being too literal about reality; on the other it is excoriated for not understanding that reality does not exist anyway.
There is, I would argue, not just a "grammar" of narrative convention, but a grammar of life—those elements without which human activity no longer looks recognisable, and without which fiction no longer seems human. WJ Harvey, following Kant, long ago proposed the notion of a "constitutive category," something which "though not in itself often the object of experience, is inherent in everything we do actually experience… without it life would be random and chaotic." The four elements of this category are, he suggests, time, identity, causality and freedom. I would add mind, or consciousness. Any fiction that lacked all five elements would probably have little power to move us. The defence of this idea of mimesis should not harden into a narrow aesthetic, for it ought to be large enough to connect Shakespeare's dramatic mimesis, say, with, Dickens's novelistic mimesis, or Dostoevsky's melodramatic mimesis with Muriel Spark's satiric mimesis, or Pushkin's poetic mimesis with Platonov's lyrical mimesis.
In one of Barthes's most quoted essays, "The Reality Effect," he ponders the question of seemingly "irrelevant" or "useless" detail in two passages by Flaubert and Jules Michelet. Barthes was a great critic in part because he thought very like a writer, or rather, thought like a writer alienated from instinct, and was drawn, like a larcenous banker, to raid again and again the very source that sustained him—literary style. Apparently "irrelevant" detail is indeed one of the great instinctive conventions of fiction-making (and not only of realism), something writers are far better at creating than explaining.
Barthes discusses Flaubert's description of Mme Aubain's room in "A Simple Heart": "Eight mahogany chairs were lined up against the white-painted wainscoting, and under the barometer stood an old piano loaded with a pyramid of boxes and cartons." The piano, Barthes suggests, is there to suggest bourgeois status, the boxes and cartons perhaps to suggest disorder. But the barometer denotes nothing; it is an object "neither incongruous nor significant"; it is apparently "irrelevant." Its business is to denote reality, it is there to create the effect, the atmosphere of the real. It simply says: "I am the real." An object like the barometer, Barthes continue, is supposed to denote the real, but in fact all it does is signify it. Realism in general, it is implied, is just such a business of false denotation
Surely Barthes is too quick to decide what is relevant and irrelevant detail. Or rather, if the barometer exists only to proclaim the real, why don't the piano and boxes, too? As AD Nuttall puts it in his excellent book A New Mimesis (1983), the barometer doesn't say "I am the real" so much as "Am I not just the sort of thing you would find in such a house?" It is neither incongruous nor significant precisely because it is merely typical. The barometer, in other words, functions much as the piano does: it creates the effect of plausibility. I would argue that realism is entirely comprised of such irrelevant/relevant detail. Secondly, one can surely accept Barthes's stylistic proviso without accepting his epistemological caveat: fictional reality is indeed made up of such "effects," but 19th-century houses like Mme Aubain's in France did contain pianos, barometers and piles of boxes. Realism can be an effect and still be true. It is only Barthes's sensitive, murderous hostility to realism that insists on this false division.
And it is clearly easier for Barthes to write about things than people. Take Orwell's famous essay "A Hanging," in which he watches the condemned man, on his way to the gallows, swerve to avoid a puddle. This, for Orwell, represents precisely what he calls the "mystery" of the life that is about to be taken: when there is no good reason for it, the condemned man is still thinking about keeping his shoes clean. It is an "irrelevant" act. Now suppose this were not an essay but a piece of fiction. The avoidance of the puddle would be precisely the kind of superb detail that, say, Tolstoy might flourish; indeed War and Peace has an execution scene very close in spirit to Orwell's essay. The avoidance of the puddle is what might be called an irrelevant detail. It isn't explicable; were it fictive, it would be there to denote precisely the inexplicable. This would be, for sure, one of the "effects" of realism, of style; we can imagine Graham Greene noticing such a thing in a novel. But Orwell's essay, because it records an actual occurrence, shows us that such fictional effects have something to tell us about the "irrelevance" of reality.
Surely realism, seen in the largest sense, is not a set of stylistic conventions but an impulse that begins with narrative itself. In Shakespeare or Defoe or Austen there is "an accurate relation to the real"—to use Peter Brook's phrase—as well as a realist's tact with detail. Coetzee, in Elizabeth Costello, has this to say about Defoe: "The blue costume, the greasy hair, are details, signs of a moderate realism. Supply the particulars, allow the significations to emerge of themselves. A procedure pioneered by Daniel Defoe. Robinson Crusoe, cast up on the beach, looks around for his shipmates. But there are none. 'I never saw them afterwards, or any sign of them,' says he, 'except three of their hats, one cap, and two shoes that were not fellows.' Two shoes, not fellows: by not being fellows, the shoes have ceased to be footwear and become proofs of death, torn by the foaming seas off the feet of drowning men and tossed ashore. No large words, no despair, just hats and caps and shoes."
Henry James saw this; it is why he wanted the novel to do everything: to be as stylistically perfect as Flaubert or Turgenev, as socially dense as Balzac, and as humanely invested in the moral life as Eliot. James found Flaubert's realism exemplary but lacking, because he felt that it did not extend to subtle, moral scrutiny of the self.
The reality of the self seems the best way to defend what is deepest and most enduring in realism; and it would connect James to Woolf, who was a much more Jamesian novelist than he thought she was, and James to Austen, a writer he resembles far more than he liked to admit. Once realism is opened up, so that it becomes a way of writing deeply about the self—a plunging into character— it comes to seem not a tradition, not a genre, but the broad central language of the novel, indeed of drama: what James in What Maisie Knew calls "the firm ground of fiction, through which indeed there curled the blue river of truth." One way of thinking about the centrality of this broad blue river is to imagine a world in which the only novel available was, say, Pynchon's Vineland and books like it. It would be a hysterical and falsifying monotony. By contrast, a world in which the only available novel was, say, Naipaul's A House for Mr Biswas would be a fearfully honest, comic, tragic, compassionate and deeply human place.
Realism, seen broadly, cannot be a genre; instead, it makes other forms of fiction seem like subgenres. For realism teaches everyone else; it schools its own truants: it is what allows magical realism, hysterical realism, fantasy, science fiction, even thrillers, to exist. Realism is nothing like as naive as its many opponents charge; almost all the great 20th-century realist novels also reflect on their own making, and are full of artifice. But it may be an impossible ideal, whereby the writer has to act as if the available novelistic methods are continually about to turn into mere convention—into "realism" as Roland Barthes or even as Rick Moody would understand it—and so has to try to outwit that inevitable ageing. The realist writer, that free servant of life, is one who must always be acting as if life were a category beyond anything the novel had yet grasped; as if life itself were always on the verge of becoming conventional.
