James Wood   How Fiction Works.
Farrar, Straus & Giroux, 252 pages, $24

Admirers of James Wood’s essay collections The Broken Estate (1999) and The Irresponsible Self (2004) will know that he has a superb eye for detail and a pithy style, and readers of his novel The Book Against God (2003) will know that he can sustain our attention through a narrative. How Fiction Works offers itself as a primer of technique and also as an enquiry into the nature of realism. The book is divided into numbered sections, which proceed irrespective of the chapters on character, dialogue, perspective and so on, and suggest that it has more of a unity than is actually the case. Wood admits that the division into chapters is slightly artificial, since all these matters tend to merge into one another. As a result, although we are frequently arrested by a virtuoso passage of analysis, the book tends to become an anthology of such passages rather than a developing argument; we miss the grip and bite of the earlier essays, as well as their greater attention to rhythm and structure.

For Wood, modern fiction begins with Flaubert. That is already a contentious statement, for, plainly, modern English fiction does not, but Wood moves from English to American, to French, to German, to Russian fiction indiscriminately, and rarely recognizes the perils of using translations. Fiction, apparently, is fiction—it isn’t even always novels, since some examples are drawn from plays and films. To choose Flaubert as the starting-point is to dismiss the crucial inheritance of the English novel from medieval romance and Elizabethan prose narrative. If we are allowed to include other genres, I would be prepared to say that Chaucer knew as much about fiction as Flaubert. Flaubert’s great innovation, in Wood’s view, was the elaboration of free indirect style to the point where the narrator/author distinction disappears and we are reading apparently artless reportage which is in fact controlled by a master stylist—above all, a master grammarian, whose habitual use of the imperfect tense in L’Education sentimentale exploits clashing “time signatures,” eliding the difference between moments of perception and larger stretches of time. Wood remarks that Proust admired this device; he did, but he also described Flaubert as a writer “whom I do not much like,” lacking in humor or sensibility, and he lamented the passiveness of the characters in L’Education sentimentale. The latter point was also forcefully made by Henry James in his two essays on Flaubert (1893 and 1902); he found Frédéric Moreau and Mme. Arnoux inadequately realized for the burden of significance they are called upon to carry. Wood’s own earlier essay on Flaubert in The Broken Estate was a sharper, and in my view more balanced, estimate of Flaubert’s strengths and weaknesses.

Since novels “work” in reaction to previous novels as well as by internal mechanisms, a chronological approach has its uses, though I would agree one does not want yet another cataloguing, pseudo-Darwinian “History of the Novel.” Wood shuttles back and forth between periods; everything seems to happen in an eternal present. A good example is the chapter rather portentously entitled “A Brief History of Consciousness.” This takes us, in thirty pages, from the Old Testament story of David to Macbeth (1606) to Crime and Punishment (1866), thence to Diderot’s Le neveu de Rameau (1784), Stendhal’s Le Rouge et le Noir (1830), back to Dostoyevsky, then to A la recherche du temps perdu (1913–1927), then to critical comments by Virginia Woolf on Dostoyesky and Ford Madox Ford on Conrad. The connection with the chapter title seems to be that later novelists discovered how to give a sense of the psychological hinterland of a character, but clearly that is not all that might be covered by “consciousness,” and, even if it were, the “stunning technical progression” of which Wood speaks cannot be shown by his grasshopper approach.

On that vexed and somewhat tedious topic of “realism,” however, Wood has helpful things to say. “Realism is not realistic,” he points out, but merely another convention, and conventions tend to become conventional; it is too often the first resort of weary hacks desperate to convince us not that “this is real” but “this is what reality in a novel like this looks like.” Realistic detail sells novels in a market dominated by readers who don’t want to think, and comes in handy for the TV or film version on which some writers have their eye all along. All this is well said, but when Wood proposes to replace the criterion of realism with one of truthfulness he obviously invites a host of objections, among them that “truth” in fiction is just as much a matter of artifice as “realism.” The only thing that matters in the end is whether we assented to the fiction while reading it. No amount of realism will be any use if we didn’t, and no amount of unrealism will be any hindrance if we did. A useful example here is Laurence Sterne’s Tristram Shandy, in which the manipulation of perspective and temporal frames of reference works simultaneously to convince us that Sterne is making the story up before our eyes, mimicking the randomness of actual perception and memory, and to remind us that the novel is an elaborate contrivance for just this purpose. The artificiality is absolutely real. I am astonished that Sterne’s novel is not mentioned at all by Wood, since it is central to every topic he discusses, but he explains that he has taken examples only from books he actually owns. All I can say is, he should buy a copy of Shandy at once.

Until very recently, a central claim for the novel was that it was a vehicle for moral and ethical enquiry. Cheeringly, Wood does not agree with the superior persons who assure us that such a view is hopelessly naïve, and explain that no novel is about anything except the act of its own composition. (That, of course, is all that many of those who have learned to write novels in creative writing courses can write about.) Wood sees in the novel the virtue that Bernard Williams found absent from much moral philosophy, that of reflecting the choice between conflicting goods rather than between a polarized good and bad. Novels should not be propaganda on behalf of a particular moral code—Wood justly deplores the “contagion of moralizing niceness” endemic in online reader reviews —but they have characteristically enquired into the sustainability of such codes (and this is true even of novelists, such as Robbe-Grillet, who disown such an agenda). A long list of examples could be given; in fact the difficulty would be who to omit.

Wood himself, in The Book Against God, examines the question of religious faith, and more broadly of the kind of faith one can put in other people—and oneself. His childhood, we discover from The Broken Estate, was divided between an evangelical home and a chorister education; he has shed his faith, not wholly gladly, but he is wise enough not to believe in fiction as a substitute, as Arnold hankered to believe in poetry. Fiction has its own kinds of truth, but its status as a provisional report on experience has been its great virtue. In Austen, George Eliot, James, and Lawrence, as in Tolstoy, Kafka, or Camus, what awaits us is not escapism or reverie but an encounter with a world often too painful to bear. Perhaps the greatest claim we can make for novels is that they show us how human beings work.