On Voice

 Originally Appeared in: Novel: 33:3 | Published: Summer, 2000

I intend in this essay on “voice” in the novel to sharpen the difference between narrative theory and novel theory as antagonistic genres of criticism. Their antagonism is not of course absolute. The development of literary criticism since the early twentieth century is the product of the cross-fertilization as well as the conflict between formalism and anti-formalism. Formalist narrative theory has often enriched nonformalist novel theory. Structuralism not only gave birth to narratology but also profoundly affected novel theorists, for example, Fredric Jameson in Marxism and Form and The Political Unconscious, Franco Moretti in The Way of the World, and Nancy Armstrong in Desire and Domestic Fiction. Other theorists have sought to synthesize rhetorical or linguistic formalisms into projects in novel theory, from Wayne Booth’s Rhetoric of Fiction to Roland Barthes’s S/Z. Indeed, the neglect of form in more recent criticism is a result in part of the fact that no new formalism has emerged since structuralism to challenge those trends which, while opening new perspectives on the politics and cultural contexts of literature, have drifted more and more into thematic, even allegorical criticism. A new formalism would be welcome today-for its insights and as something to contend against.

The principal strength of novel theory lies in its tendency to conceive the novel as a specific cultural and literary form, as when Moretti analyzes the form of the Bildungsroman from Goethe to Flaubert in cultural rather than formalist categories. His thesis is that as “Europe plunges into modernity, but without possessing a culture of modernity,” it finds in “youth a “material sign” to represent the “spiritual content” of modernity (5). “[Tlhe new and destabilizing forces of capitalism impose a hitherto unknown mobility,” but also give “rise to unexpected hopes, thereby generating an interiority not only fuller than before, but also . .. perennially dissatisfied and restless” (4). Youth, shed of many of its actual features, becomes a culturally potent symbol capable of designating at once this mobility and this restless, dissatisfied interiority. “If youth, therefore, achieves its symbolic centrality, and the ‘great narrative’ of the Bildungsroman comes into being, this is because Europe has to attach a meaning, not so much to youth, as to rnodernify …. Youth is, so to speak, modernity’s ‘essence,’ the sign of a world that seeks its meaning in thefuture rather than in the past” (5). The cultural symbolization of modernity as youth generates the formal possibilities of the Bildungsroman, but at the same time it creates the genre’s formal limit and inner contradiction: the brevity of youth ‘yorces the a priori establishment of a formal constraint” (the story cannot proceed beyond the protagonist’s maturation); consequently, the Bildungsroman’s means of representing modernity tends to betray the “intrinsically boundless dynamism” of the modernity it is called upon to represent in the first place (6). To examine how the different trends within the Bildungsroman cope with this latent contradiction in the form, Moretti foregrounds the problem of plot and draws extensively on analytic models developed by narrative theory and narratology. His work remains within the critical tradition of novel theory, however, because its formal-structural analysis of plot strategies is ancillary to the social-symbolic determinations of the Bildungsroman’s form, origins, inner contradictions, cultural value, and aesthetic effects.

Novel theory diverges sharply from narrative theory when it comes to understanding the novel’s relation to the folktale and other popular fiction, such as romances or detective novels. For narrative theory, all these narrative forms have the same, or at least homologous, elements and structuring principles. The novel is a variant of narrative in general. Not so, in the eyes of novel theory. Whereas structuralism takes Propp’s Morphology of the Folk Tale as its inaugural model for a structural analysis of all narrative, including the novel, Bakhtin considers the folktale a “primary genre” that the novel reinscribes-rewrites and reaccentuates- according to its own specifically novelistic imperatives and purposes. As for popular fiction-and by extension other narrative forms that emerge alongside the history of the modem novel, like cinema-Bakhtin and the early Lukacs insist that these popular and mass-cultural forms are derivatives and artistic reductions of the novel form itself, not equivalent or parallel developments. Lukacs makes the following assertion in the context of assessing why critics in the early twentieth century failed to grasp the aesthetic importance of the novel compared to genres like epic, poetry, and tragedy whose origins lay in the premodern world:

The novel, in contrast to other genres whose existence resides within the finished form, appears as something in process of becoming. That is why, from the artistic viewpoint, the novel is the most hazardous genre, and why it has been described as only half an art by many who equate having a problematic with being problematic. The description may seem convincing because the novel-unlike other genres-has a caricatural twin almost indistinguishable from itself in all inessential formal characteristics: the entertainment novel, which has all the outward features of the novel but which, in essence, is bound to nothing and based on nothing, i.e. is entirely meaningless.. . . [Superficial likeness can almost lead to the caricature being mistaken for the real thing. But a closer look will always, in any concrete case, reveal the caricature for what it is. (72-73)

Bakhtin, who cuts a much wider swathe than Lukacs in defining the novel and views folkloric and popular culture as a source of the novel’s aesthetic, holds much the same view of the relation of the novel to its derivatives and caricature. Valuation is integral to novel theory’s understanding of literary forms. By contrast, narrative theory and narratology aspire to a value-neutral conception of narrative; indeed, going back to the seminal issues of the French journal Communications, 8 and 11, structuralist narrative theory gaily combines Raymond Roussel and Ian Fleming, La Princesse de Cleves and France-Soir, Flaubert and Bororo myth, to banish aesthetic value from narrative analysis.

The tendency of novel theory to define the novel as a genre that is sui generis does lead to certain impasses and quandaries, whether in the theory of the novel as such or of a subgenre like the Bildungsroman. The novel theorist ties the interpretation of the literary form to a particular historical and philosophical understanding of the salient features of modern society and culture; inevitably, the evaluation of the form has to exclude or devalue certain works and tendencies. The Theory of the Novel concludes with a reflection on Dostoevsky as a limit-case which cannot be comprehended within Luk6csrs own conception of modernity and the novel: “he, and the form he created, lie outside the scope of this book. Dostoevsky did not write novels … [!I” (152). For Lukdcs, Dostoevsky was strictly speaking uninterpretable, an incomprehensible dead end within the modem world or a harbinger of something new that only “later artists will one day weave into a great unity” (153). Moretti delineates the complexities of the Bildungsroman with extraordinary subtlety and precision, and yet, because of the very understanding of history and culture that his reading unfolds, is led to disqualify the English tradition, from Austen and the Brontes to Dickens, as failed or flawed attempts to realize the genre’s possibilities. His valuations, inseparable from his historical-cultural interpretation of the genre’s form, open onto controversy as dramatically as Lukdcs’s exclusion of Dostoevsky. Bakhtin’s placing of Dostoevsky at the very heart of the modern novel or feminist critics’ revaluation of English courtship and marriage novels suggest the sorts of controversy to which novel theory is intrinsically prone. Narrative theory is spared such difficulties. In construing its object as narrative structures, strategies, and functions without regard to the kind of discourse in which they occur, it achieves a self-consistent critical language unencumbered by the need to assess either the aesthetic or the social value of the novel or to account for its contradictory imperatives. The self-consistency of narrative theory is not, however, a point in its favor. On the contrary. For isn’t the moment one’s thought begins to match up perfectly to its object the exact moment to doubt its truthfulness? For me, an indelible lesson taught by Paul de Man is to recognize that it is in the nature of literary works and interpretation that every genuine critical understanding arrives at its maximum insight only to disclose its ineradicable blindspot.


The general antagonism between novel theory and narrative theory takes on a new urgency in light of the tendency in several recent novelists to cast the narrative voice as the writer, not a narrator in the sense of a storyteller or observer or presiding consciousness. In their work, narration is an act of writing, not storytelling. That a novel is a product of writing is hardly controversial. It is a selfevident fact known to all readers. But novelists who foreground the act of writing seem to challenge a set of well-entrenched narrative conventions and reading habits which have been variously codified in twentieth-century criticism.

According to those conventions, habits, and codes, narration takes place in an imaginary space. The first-person narrator, even if a participant in the story, recounts it from a fictive place unlocated in concrete space and time. Where is Pip when he recounts his encounter with the convict, his days at Miss Havisham’s, or his arrival in London? The question is superfluous, indeed silly. In third-person narrations, from Madame Bovary to Portrait of the Artist as a Young Man, the supposed narrator occupies an imaginary space in the simple sense that he does not “exist,” either on the plane of reality of the story or that of the book. This creation of an imaginary space of narration is a complex stylization, a kind of rhetorical zone in which the narrator “recounts” events-actions, emotions, thoughts-as though he or she has “observed” them, though no such space of witness exists within or outside the story told. That we accept this rhetoric of recounting and observing, this imaginary space from which someone who is no one addresses us, is at the very least a significant achievement in our modern capacity for alienation.

For that very reason, the novel form bugged Walter Benjamin, who saw the rise of the novel as the death of genuine storytelling and, with it, of the very capacity to communicate one’s experiences and wisdom, the kind of experience and wisdom that accumulated, according to his archetypes, in the lives of sailors, farmers, travelers, and urban craftsmen. Modem critics have on the whole, unlike Benjamin, embraced the alienations of the novel, but then set about to describe and codify narrative conventions so as to tame these alienations into structured unities in the name of various values: aesthetic mastery, impersonality, objectivity, irony, judgment, controlling consciousness, and so on.

Our whole conceptualization of narrator and implied author has generated a rich vocabulary for clarifying narrational structures, but there is something amiss in it. For it ultimately neglects the specificity of novelistic writing. An implied author occupies a perspective; an author engages in an act of writing. Narrative theory in effect reduces the act of writing, that is, the actual author’s practice, to nothing more than the process of creating the implied author, who is then stationed as the outer limit of the narrative theorist’s interpretive and analytical attention. What then is at stake when contemporary novelists reverse this scheme and identify narration with writing, evacuating or overrunning that imaginary space in which we read-or hear-an implied author or narrator addressing us? The purposes and effects of novelists’ foregrounding of writing are extremely varied, as are the stylistic and formal challenges posed by their active interrogation of the imaginary space of narration.


Salman Rushdie writes a Menippean voice in Shame, riding roughshod over tropes of invention and tropes of representation: “The country in this story is not Pakistan, or not quite …. My story, my fictional country exist, like myself, at a slight angle to reality” (23-24). He opens avenues of commentary with essayistic asides on politics and history and mulls over the plot choices he makes and the nuances of the theme of shame. At one point, he tells how a character in the novel was inspired by the story of a Pakistani girl in London murdered by her father for “making love to a white boy”:

My Sufiya Zinobia grew out of the corpse of that murdered girl …. I even went so far as to give the dead girl a name: Anahita Muhammad, known as Anna …. But finally she eluded me, she became a ghost, and I realized that to write about her, about shame, I would have to go back East, to let the idea breathe its favourite air. Anna, deported, repatriated to a country she had never seen, caught brain-fever and turned into a sort of idiot.

Why did I do that to her? (124-25)

The mixing of fantastic and realistic modes in Rushdie, the ragged edges he makes between invention and representation, has been generally described as magical realism or metafiction. I think both descriptions are misleading, though the influence of Gabriel Garcia Mdrquez or Thomas Pynchon on Rushdie is undeniable. Garcia MArquez’s magical realism works by making the connection between the fantastic and the realistic seamless; he narrates utterly implausible and plausible events alike in the style of verisimilitude. He creates the unified sense of a world which is neither mundane nor fanciful because both at once. Rushdie, by contrast, spins fantastical episodes as imaginative and comic commentary on social realities. It is in that sense that he revives Menippean satire within contemporary fiction. By the same token, he does not write metafiction in the sense contemporary criticism gives the term. The passage about the creation of Sufiya Zinobia does not draw you into an epistemological paradox about the possibilities of knowing or representing reality. It jarringly layers the real-world sources of the story and its elaboration into a full-blown fictional tale and signals the yet more complex historical referent of that tale. That the London news story the writer says he read may itself be a fabrication is moot. In foregrounding the act of writing, with its heterogenous sources and referents, Rushdie is recovering the Menippean tradition which, through Rabelais, Sterne, and Dostoevsky, has shadowed other developments in the history of the novel.

Rushdie exemplifies the tendency in twentieth-century fiction that Milan Kundera identifies as a spiraling back to formal possibilities of “the nearly forgotten aesthetic of the novel previous to the nineteenth century” (Testaments Betrayed 74). In Rushdie’s recovery of Menippean satire-with its mixture of fantasy and reality, comic exaggeration, intellectual parody, social commentary-the voice of the novel emerges as the writer’s creation, putting his intentions and purposes openly at stake in the narration itself.


Christa Wolf’s novel Patterns of Childhood is unmistakably autobiographical, but it is not narrated autobiographically. It alternates among three distinct time frames and different voices. The main story is presented in the third person as the story of a young girl named Nelly Jordan, her parents, and her younger brother Lutz. Nelly, like Christa Wolf, is born in 1929 and hence four years old when the Nazis come to power and sixteen at the end of the war. She lives in the German town of L. The second time frame is the summer of 1971 during a trip the writer takes to her birthplace in the Polish town of G., formerly the German town of L. She is accompanied by her husband and daughter and her younger brother Lutz. This narrative is told in the second person in the form of a memoir the writer addresses to herself. The third strand is likewise in the second person and is presented as the diary the writer keeps during the composition of the novel, from 1972-1974.

The project first arose years earlier when Frankfurt students avidly following the Auschwitz trial disturbed the writer’s assumption that only the older generation of Germans had secrets from the Nazi era: “You were unprepared for their demand that you yield your secret …. As though you could be relieved of the duty to lay a hand upon your own childhood” (248). But she was unable to begin the writing until she recast the story that was emerging from her own remembering as someone else’s life story. Hence third-person narration: “Gradually, as the months went by, the dilemma crystallized: to remain speechless, or else live in the third person” (3). Remaining speechless was, she says, impossible, living in the third person merely strange.

Next she planned a novel written completely in the third person, presumably shuttling between a woman’s trip in 1971 to her birthplace and the story of her childhood there during the Third Reich, minus the writer’s diary. “The sudden switch from the third to the second person (which only seems to be closer to the first)” came “the morning after a vivid dream” (118) in which she discovers a mysterious object and tricks someone into believing that it is the thing she had been entrusted with and was then accused of stealing; by finding this mysterious object, she proves she had merely forgotten about it: “a true memory lapse!” (119). The writer takes the dream as a warning against her assumption that the child whose experience the adult seeks to remember and the adult herself must be, or can become, the same. Some discoveries, like the mysterious object in the dream, have the feel of memory lapses but are in fact hoaxes that merely exonerate the one who “remembers.” So, just as the writer is stymied in speechlessness in the face of first-person narration, she now rejects the purely third-person narration because it presumes an identity between child and adult and does not guard the writer herself against turning memories and memory lapses into a hoax. Therein lies the dilemma of Wolf’s search for Nelly Jordan. To find her she must sharpen the estrangement: “That’s when you had to realize that you could never again be her ally, that you were an intruding stranger pursuing not a more or less well-marked trail but actually the child herself” (119).

The therapeutic expectation of modern autobiography is jeopardized in Patterns of Childhood. Not only does autobiography suppose the power to recover forgotten traumas, suppressed desire, and the oldest misunderstandings, but the very process of recovering them is supposed to secure and deepen one’s identity. As Wolf’s novel delves into the adolescent Nelly’s relationship to Dr. Julianne Strauch-her inspiring teacher, role model, love object, and a devoted Nazi-the crisis of the writer’s relation to her character reaches a new extreme. What the writer’s memory recovers is how thoroughly Nelly’s schooling in the values that the writer now most prizes within herself, independence and self-respect as a woman, come from her devotion to Julianne and enthusiastic participation in Hitler Youth. The more the writer discovers the entanglement of Nazism and everyday life, ideology and identity, the less her life returns to her as her own. There is no bridge between her past and present: “The closer she gets to you in time, the less familiar she becomes” (211). The novelistic strategy of estrangement, originally undertaken through the splitting of voices in order to overcome speechlessness and bad faith, does not open any dialectic path to overcoming a lived estrangement; indeed, the writer’s estrangement from her own elaborately disclosed past has found form in this stylistic experiment but not a healing transcendence. In the end, she asks, “And the past, which can still split the first person into the second and the third-has its hegemony been broken? Will the voices be still?” She answers, “I don’t know” (406).


The voice of Toni Morrison’s Jazz is an I-narrator who has no role in the story, nor even an identity, though her verbal style, recollections, and familiarity with events and surroundings place her in the same Harlem of the 1920s as the characters. She belongs to the neighborhood, but her omniscience exceeds what even the most inveterate busybody could acquire in gossip and talk. And yet nothing in the text raises questions about her reliability in the conventional sense.

When the final chapter uncovers a turbulence in the relation between the voice and the characters, it also makes evident something that is palpable throughout the novel: the novel’s voice is an amalgam of the unidentified neighbor and the writer. Writer, not “implied author,” because the chapter explores the making of the characters and story not merely the “observation” and “recounting” of them: “Pain. I seem to have an affection, a kind of sweettooth for it …. I break lives to prove I can mend them back again” (219). The difference between the neighbor and the writer is not a structured separation of “narrator” and “implied author,” for they confront the same crisis when faced with the story’s outcome, and they articulate the same sentences, share the same voice. Their difference is more akin to the two sides of a Mobius strip: the neighbor is a gossip and chronicler who embellishes what she knows about the protagonists up to the crisis point where their actions throw in doubt her speculations about their inner reality and their fate; the writer is a creator who has made characters who, against her own expectation, resist the fate to which the story seems to destine them. Thus, neighbor and writer admit (in one voice) to misunderstanding Joe and Violet: “I was sure one would kill the other” (220).

The moral pattern inscribed in the story of Joe and Violet Trace suggests an unexpiated violence that inevitably spawns more violence. Joe’s murder of his young lover Dorcas is never comprehended or forgiven, and Violet’s mad attempt to disfigure the dead girl’s face is followed by her yet madder act of setting Dorcas’s photograph on the mantle in an uncanny homage meant to remind Joe of what he can never undo and ambiguously give Violet herself an image of the daughter she lost long ago in miscarriage, a daughter Joe never wanted. Neighbor and writer have read in these bluesy actions auguries of recurrent violence: “So I missed it altogether. I was so sure one would kill the other. I waited for it so I could describe it. I was so sure it would happen. That the past was an abused record with no choice but to repeat itself at the crack and no power on earth could lift the arm that held the needle” (220). The moral inevitability of the agon between Joe and Violet is thwarted, as this blues man and blues woman do an improvisation of their own, eluding the outcome that the novel’s voice was prepared for: “I was so sure, and they danced and walked all over me. Busy, they were, busy being original, complicated, changeable-human, I guess you’d say, while I was the predictable one, confused in my solitude into arrogance, thinking my space, my view was the only one that was or that mattered” (220).

What is this solitude and arrogance, this space and view, if not the solitary space and arrogant view of omniscience? Morrison is interrogating what cannot be known of another-whether a character in a novel or the troubled couple across the street-just as Christa Wolf is interrogating what cannot be known of oneself. “I got so aroused while meddling, while finger-shaping, I overreached and missed the obvious. I was watching the streets, thrilled by the buildings pressing and pressed by stone; so glad to be looking out and in on things I dismissed what went on in heart-pockets closed to me” (Morrison 220-21).

There is no novel without omniscience, yet every omniscience is limited; therefore, there is no omniscience. Our readerly obsession with authorial judgments, psychological and moral, ideological and political, easily misses this paradox inherent in novel writing. For to create a character is also to create the heart-pockets closed to writer and reader alike. Morrison is not simply playing a narratological game here, though it would be easy to describe it as such by claiming that she has her narrator marvel at the mystery of the characters in order to “motivate” or rationalize an ending in which those characters slip the yoke that the story’s logic enticed the reader into expecting.

The mystery of character, however, belongs to novelistic truth rather than narratological lie, and novelists frequently remark on their sense of a character’s relative opacity and separateness from them. “My characters always begin by being an enigma to me,” says Nadine Gordimer. “I know something about them. And then as the novel, the story, develops I learn more. I don’t know how. I don’t have a preconceived idea of everything that they are” (“Family Plots” 21). Her novels frequently internalize this always incomplete probing by presenting the main character obliquely-Rosa in Burger’s Daughter, Hillela in A Sport of Nature, Duncan in The House Gun-as a puzzle the other characters struggle to comprehend. Characters also recoil on their authors in the wake of writing; when Elias Canetti finished Auto-da-F&, he fell into confused remorse and guilt for inventing the death by fire that was his protagonist Kien’s fate. The limits or gaps in omniscience are in fact part of the very shape of the making of characters; the precise nature of the limit acquires significance within a novel’s larger patterns and purposes. Remarking on the “sudden illumination” that marks “the metamorphoses of Tolstoy’s characters,” Kundera writes, “Pierre Bezukhov is transformed from an atheist into a believer with astonishing ease. All it takes is for him to be shaken up by the break with his wife and to encounter at a post house a traveling Freemason who talks to him. That ease is not due to lightweight capriciousness. Rather, it shows us that visible change was prepared by a hidden, unconscious process, which suddenly bursts into broad daylight” (Testaments Betrayed 215).

In Jazz, the dimension of Joe and Violet that the story’s moral design could not bring to light is designated “Something rogue.” Their survival together entails “Something rogue. Something else you have to figure in before you can figure it out” (228). The narrative crisis dramatized in the final chapter touches on the cultural and historical project Morrison poses for herself in writing Jazz. The novel’s larger project is to create a vision of the African American lifeworld of Harlem in the 1920s. The architecture and rhythm of city life, the historical awareness of black migrants, war veterans, and riot victims, and the constant presence of jazz in the culture of everyday life are the rich and rugged components of this represented lifeworld.

In Joe and Violet’s story of love and violence, craziness and survival, Morrison sets out to render a blues story novelistically. There is a rift, however, between novelistic design and the blues aesthetic. “It never occurred to me that they were . . . putting their lives together in ways I never dreamed of” (221); their survival is an improvised dance: “they danced and walked all over me” (220). The voice Morrison stylizes in Jazz-the neighbor busying herself with other people’s pain and the writer in search of a historical lifeworld that beckons just beyond the reach of her own experience-this voice posits her own understanding of the blues and sees it exceeded by what’s unseen: “I started out believing that life was made just so the world would have some way to think about itself, but that it had gone awry with humans because flesh, pinioned by misery, hangs on to it with pleasure …. I don’t believe that anymore. Something is missing there. Something rogue. Something else you have to figure in before you can figure it out” (227- 28).


Two originators of novel theory, the early Lukacs and Bakhtin, do not presuppose an imaginary space of narration. They look at narration from the standpoint of writing. Rather than treating the novel as a type of narrative among others (myth, folktale, film, and so on) as narrative theory and narratology do, the novel theorists consider the novel a specific, though diverse and polyglot, cultural form and social practice. Its aesthetic is incommensurable with that of other narrative arts.

Luk6cs approaches the act of novelistic writing in his densely dialectic account of irony in The Theory of the Novel. I will quote the central passage and then gloss it:

[Irony in the novel] signifies an interior diversion of the normatively creative subject into a subjectivity as interiority, which opposes power complexes that are alien to it and which strives to imprint the contents of its longing upon the alien world, and a subjectivity which sees through the abstract, and therefore, limited nature of the mutually alien worlds of subject and object, understands these worlds by seeing their limitations as necessary conditions of their existence and, by thus seeing through them, allows the duality of the world to subsist. At the same time the creative subjectivity glimpses a unified world in the mutual relativity of elements essentially alien to one another, and gives form to this world. (74-75)

To paraphrase: novelistic irony is attained as the writing subject splits in two, becoming (1) the protagonist’s inner striving and desire, whose efforts are represented in story coming up against the limits of the social world, political forces beyond his control, the will of others, and so on; and (2) the novel’s voice which, unlike the protagonist, (a) sees through the abstractness of the conflict between the latter’s inner world and the outer world; (b) understands that the outer world is necessarily meaningless because without any animating orientation and that the subject’s inner animation is necessarily blind to the way of the world; and therefore (c) negates neither inner nor outer reality in favor of the other. At the same time, the creative subject who has thus split in two in making the novel (d) glimpses and (e) gives form to this unified world of mutually antagonistic subjective and objective forces.

According to Lukacs, the subject who achieves this form-giving glimpse recognizes “the antagonistic nature of the inner and outer worlds . . . as necessary.” Who then is this subject? Certainly not a presiding consciousness, since it does not contemplatively possess these necessarily antagonistic, mutually relative worlds. And not an implied author, whether conceived as an omniscient judge of inner and outer realities or as an effect of textuality. Lukacs insists, on the contrary, that this subject “is just as empirical-just as much part of the outside world, confined in its own interiority-as the characters which have become its object” (75). This subject is the writer.

Bakhtin likewise insists on the empirical existence and worldly practice of the writer. The essential reference point of novel theory for Bakhtin is the process of composition; he conceives of the novel as discourse, an act of communication ventured in, and venturing to alter, concrete public spheres. The composition or construction of a novel takes as its raw material a variety of discourses active in society at large and re-voices them. In the terms he developed in the essay “The Problem of Speech Genres,” these pre-existing social discourses are “primary genres-for example, “the rejoinder in dialogue, everyday stories, letters, diaries, minutes, and so forth”-which “secondary, complex genres,” like novels, “play out” (Speech 98). The novel engages the sociality of communication on, as it were, two fronts: on the one hand, it incorporates into its very construction discourses originating in several social contexts, public and private; and, on the other hand, it addresses itself to, intervenes in, an actual public realm.

As part of his account of the voicing a novel acquires in this layering of discourse, Bakhtin gives a more ample and variegated account of narrational structures than any other twentieth-century critic. He details a wealth of literary “convention[s] and semi-convention[s]” through which novelists create “images of substitute authors, editors, and various kinds of narrators” (Speech 98). He does not, however, then conceptualize the novel as those Chinese boxes by which narrative theory encases story and character inside the box of narrator-narratee and then encases that box within the implied author-implied reader box, leaving the writer and the reader mysteriously outside the narrative boxes altogether. On the contrary, Bakhtin insists that “the most complex and ultra-composite work of a secondary genre”-like a novel-“as a whole (viewed as a whole) is a single integrated real utterance that has a real author and real addressees whom this author perceives and imagines” (98-99).

Bakhtin does not attribute to the real author anything like sovereignty over the discourse he or she produces. Like Lukacs, though with a significantly different analysis of the novel form, he locates the writer’s subjectivity in the social practice of writing. On the one hand, the author is “the creator of the work itself, although he is located outside the chronotypes represented in his work, he is as it were tangential to them. We meet him (that is, we sense his activity) most of all in the composition of the work (Dialogic 254); and on the other hand, the author is present in the “accentuations,” “accents,” “refracted” intentions, with which he or she inflects the contentious dialogical play of the social discourses within the work.

While Bakhtin locates the writer’s empirical existence in the composition and communicative action of the novel within what Hannah Arendt calls the worldly space of the public realm, Lukgcs locates it in the practice of novelistic irony. As creator, the writer puts his or her own subjectivity in play by projecting it into the interiority of the character enmeshed in the social world represented in the novel. As observer, the writer attains the irony by which he or she “sees through,” “understands,” “recognizes,” “glimpses,” and “gives form” to the relation of protagonist and society, inner and outer reality. As creator, the writer draws on his or her own inner reality, fully as finite and enmeshed in outer reality as the character’s; as observer, the writer’s understandings are as worldly and fragile as any act of understanding. In Lukacs words: “In the novel the subject, as observer and creator, is compelled by irony to apply its recognition of the world to itself and to treat itself, like its own creatures, as a free object of free irony” (75).

Neither Luk6cs nor Bakhtin brackets or boxes out the real author in accounting for the form of the novel. Therein lies another crucial difference between novel theory and narrative theory; novel theory is as preoccupied with form as narrative theory is, but it has a nonformalist conception of form. Narrative theory’s formalism, in eschewing the real author, the empirical subject who writes, installs a semblance of authorial consciousness far more sovereign and unified than anything LukAcs or Bakhtin could fathom. Narrator-narratee and implied author-implied reader preside over homogenous worlds of representation and meaning, neatly removed from the antagonistic unity that the Lukacsian subject glimpses between inner and outer worlds and from the public world in which the Bakhtinian author inflects contentious discourses with particular values, judgments, and perspectives.

Glimpse and inflection-these terms suggestively affirm that the ultimate manifestation of the writer’s presence in the novel, whether conceived as the attainment of irony or as the communication of intention, is partial, finite, even precarious, because it dwells in the empirical world of experience and communication. The more fully the writer’s subjectivity and the actual act of novelistic writing are accounted for, the less writing and reading are idealized. Criticism needs to reaffirm this worldly space of the novel rather than the imaginary space of narrative.


Where then did the regulative, homogeneous conception of narrative voice come from? What are the sources of our critical and readerly habit of bracketing out the writer and boxing in narrators and implied authors? Like most critical concepts and terms, the current vocabulary of narrative voice is a response to specific innovations and conventions in the history of fiction itself. The languages of criticism are prepared by the language of literature.

Edgar Allan Poe’s innovations in narrative form crucially consolidated the conventions that fix the narrator in an imaginary space. His first-person narrators typically address their story to a hypothetical listener, in the sense that there is no interlocutor on the same plane of reality or representation as the narrator himself. It is this void marking the absence or sheer hypothesis of a listener that narrative theory fills, rather pointlessly in my view, with the theoretical postulate of a “narratee.” That Poe’s innovations took place in short stories rather than novels suggests how far-reaching those innovations were, how much control over the composition of prose was required to effect a structured distinction between narrator and (implied) author and uphold it through an entire narrative.

In “The Black Cat,” the narrator himself occupies a concrete space and time: he is in his jail cell on the eve of his execution. And there is the fiction that he is writing-“the most wild yet most homely narrative which I am about to pen” (223)-but, as with Poe’s other first-person narrators, there is no one to receive this narrative on the same plane of reality as the narrator.

The creation of a voice that sustains a consistent distinction between narrator and author is far more than a matter of distinct identities. Innumerable earlier narrative styles would avoid confusing the author with a narrator on the verge of death. Poe’s achievement lies in organizing every element of the story into a pattern that makes the author/narrator distinction central to story’s total aesthetic, moral, and thematic effect. Just as he advocates in his essay “The Poetic Principle” an aesthetic whose unity of effect and affect is impossible to achieve in a long poem, his prose aesthetic found its appropriate boundaries in the short story form rather than the novel. But the stylistic innovations and conventions he established through that form profoundly influenced the novel and novel criticism.

The organizing principle that structures voice in “The Black Cat” lies in the, logic governing the sequence of events, the “chain of facts” in the narrator’s words, through which the story unfolds. “I am above the weakness of seeking to establish a sequence of cause and effect, between the disaster and the atrocity,” the narrator insists. “But I am detailing a chain of facts-and wish not to leave even a possible link imperfect” (225). The narrator is the protagonist of an uncanny story, a “most wild yet most homely narrative,” but he repudiates superstition, like the belief that black cats are “witches in disguise,” to explain the happenings in his story, and he affirms that he is not mad. Yet he knows it will take “some intellect more calm, more logical, and far less excitable than my own” to see “nothing more than an ordinary succession of very natural causes and effects” in the details that fill the narrator himself with “horror” and “awe” (223). This allusion to another understanding signals the two tiers of perspective essential to the work’s ultimate effect.

The “chain of facts-this “series of mere household events” whose “consequences” have “terrified,” “tortured,” “destroyed” the narrator (223)-acquires its uncanniness as the mere events begin to repeat or allude to one another. The narrator, recounting how the “disease” of alcohol altered his personality, tells of the cruelties he increasingly directed at his wife and their several pets. In the first sequence of events, one night he turns on his favorite pet, the black cat, and “deliberately cut[s] one of its eyes from its socket” (224); some while later, in a second fit, he kills the cat by hanging it from a tree in the yard; soon thereafter, the house is destroyed by fire, and when the narrator joins the neighbors gathered around the ruins the next morning he discovers they are staring in amazement at “the figure of a gigantic cat,” noose around its neck, embedded in the house’s one remaining wall (225). This uncanny return of the cat, the sign of his crime, sets in motion the struggle of reason with conscience, science with fancy:

When I first beheld this apparition-for 1 could scarcely regard it as less-my wonder and my terror were extreme. Buf at length reflection came to my aid. The cat, 1 remembered, had been hung in a garden adjacent to the house. Upon the alarm offire, this garden had been irnmediatelyfilled by the crowd-by some one of whom the animal must have been cutfrom the tree and thrown, through an open window, into my chamber. This had probably been done with the view of arousing mefrom sleep. The falling of other walls had compressed the victim of my cruelty into the substance the freshly-spread plaster; the lime of which, with the names, and the ammonia from the carcass, had accomplished the portraiture as I saw it.

Although I thus readily accounted to my reason, f not altogether to my conscience, for the startling fact just detailed, it did not the less fail to make a deep impression upon my fancy. (225-26)

The narrator’s experience of dread begins with the second sequence of events. Wanting to replace the pet he’s killed, he purchases another black cat one night at a tavern, a cat as large as the first and resembling him except for an “indefinite splotch of white” on its breast. Next morning, however, the narrator discovers that this cat, “like Pluto, . . . also had been deprived of one of its eyes.” His dread intensifies to torment when his wife points out that the cat’s white markings are the image “of the GALLOWS!–oh, mournful and terrible engine of Horror and of Crime of Agony and of Death!” (227). Faced with these signs of his previous crimes, the narrator acquires a “hatred of all things and of all mankind,” more and more directing his fury at his wife. One day as they are going down to the cellar, the cat follows them and nearly trips him; “forgetting in my wrath the childish dread which had hitherto stayed my hand,” he picks up an axe and aims at the cat. “But this blow was arrested by the hand of my wife. Goaded by the interference into a rage more than demoniacal, I withdrew my arm from her grasp and buried the axe in her brain. She fell dead upon the spot without a groan” (228). He decides to pull out the bricks from a cellar wall that “had lately been plastered,” place the corpse within, and repair the wall. When the police come to investigate the wife’s absence a few days later, they search the entire house, including the cellar; satisfied, they are about to leave when the narrator-” in the rabid desire to say something easily-remarks on the construction of the house and raps on the wall where his wife is buried. A “voice from within the tomb” answers, “quickly swelling into one long, loud, and continuous scream, utterly anomalous and inhuman.” The police tear down the wall, revealing the corpse. “Upon its head, with red extended mouth and solitary eye of fire, sat the hideous beast whose craft had seduced me into murder, and whose informing voice had consigned me to the hangman. I had walled the monster up within the tomb” (230).

So ends the narration, the sign of the narrator’s crime once again returning to him, the howling of the walled-in cat repeating, down to the details of plaster and fire, the hanged cat’s return in the wall of the ruined house. The two sequences are a series of mere facts:

(1) cuts out cat’s eye (2) hangs cat (3) cat embedded in the wall

(4) one-eyed cat (5) gallows markings (6) kills wife (7) cat howls in wall
Through the workings of sheer chance and happenstance, these events fall into a pattern in which each crime returns in a sign of the crime; moreover, the second sequence of events reiterates the first:

(1) cuts out cat’s eye [crime] (2) hangs cat [crime] (3) cat embedded in wall [sign of crime (2)

(4) one-eyed cat (5) gallows markings (6) kills wife (7) cat howls in wall
[sign of crime (I)] [sign of crime (2)] [crime] [sign of crime (3)(6)]

The sense of fatality so important in Poe culminates in the fact that the narrator’s imminent encounter with the hangman gathers up the imagery of noose and gallows into a kind of final manifestation of the symbolic in the real
(8) He is to hang [2,3,5,6]

As in the E.T.A. Hoffmann tale Freud analyzes, Poe here produces the uncanny by projecting a set of symbolic equivalences and doublings onto the plane of the character’s reality. The symbolic happens as the real, not in the mode of a symbolically rich universe realizing itself in events as is the case with myth, fairy tale, or providential history, but in the disenchanted world of “natural causes and effects,” the “chain of facts,” “a series of mere household events.” The uncanny requires the “homely narrative” as its base.

The structured distinction between narrator and (implied) author, I want to suggest, is not primarily a matter of identity or personality. In this story it lies, rather, in the distinction between two attitudes toward the uncanny play of the symbolic and the real. For the narrator, the uncanny is an experience of dread, the moral dread of the criminal whose psychology is Poe’s principal theme and fascination. The narrator’s reason struggles almost helplessly against the feeling of a supernatural force which even at the last moment prevents him from recognizing his own agency in his crime: “the hideous beast whose craft had seduced me into murder” (230).

What then is the second, differentiated attitude, the one we now so commonly attribute to the implied author? My quarrel with the narrator/implied author formula is that it reduces the author’s activity to contemplation. Accordingly, “The Black Cat” would be construed as a story recounted by a narrator contemplated by an implied author. But to get at what’s going on, let’s restate this from the standpoint of writing rather than contemplation: the author writes the narrator telling a story. What needs to be described is this process of writing the narrator telling. Poe projects the symbolic plane of doublings, signs, repetitions, and equivalences onto the plane of the narrator’s reality. The two structurally distinct attitudes are the result. For counterposed to the dread evoked for the narrator by the interpenetration of the symbolic and the real is the aesthetic mastery of the writing itself, a mastery in which the prevalence of the symbolic over the real occurs without recourse to superstition, magic, or providence, without, that is, challenging “natural causes and effects.” The unity of effect which is the aim of Poe’s aesthetic takes the form in “The Black Cat” of the simultaneity and differentiation of moral dread and aesthetic mastery. More precisely, the dread is evoked and mastered within the aesthetic attitude, that frisson of fascination with horror that replicates in reverse the narrator’s own transfixed horror at his fate. Pw’s act of writing-the-narrator-telling is the stylization that creates this effect, transforming horror into fascination, dread into aesthetic contemplation. It is this contemplative attitude that narrative criticism has codified into the normative stance of the implied author and implied reader.


Poe’s achievement is immense, and if my hypothesis is correct that it established the conventions of the imaginary space of narration and the structured distinction of narrator and (implied) author, it has had a profound effect on reading habits and criticism. When Bakhtin surveys the long history of the novel, taking stock of its stylistic variability, he assesses the role of “posited authors,” storytellers, and first-person narrators in light of an aesthetic very different from Poe’s. “The speech of such narrators is always another’s speech (as regards the real or potential direct discourse of the author) and in another’s language (i.e., insofar as it is a particular variant of the literary language that clashes with the language of the narrator) …. All forms involving a narrator or a posited author signify to one degree or another by their presence the author’s freedom from a unitary and singular language, a freedom connected with the relativity of literary and language systems; such forms open up the possibility of never having to define oneself in language, the possibility of translating one’s own intentions from one linguistic system to another” (Dialogic 313-15).

Bakhtin thus sees in the flexibility and relativity of “double-voiced” narrations the writer’s leeway to experiment with his or her commitment to the norms and meanings of a particular discourse. Keying on the resources that such “a refracting of authorial intentions” affords the comic novel in particular, he stresses the “variety of different distances between distinct aspects of the narrator’s language and the author’s language” (315). Poe’s innovative stylization turns the relativity of double voice to a more regulated, unified purpose. He renders it “monological” in Bakhtin’s terms or “parsimoniously plural” to borrow Roland Barthes’s term in S/Z. The underlying discourses of confession, Gothic tale, psychology, and criminology are blended into the two-tiered attitude that fixes the author’s relation to the narrator as contemplative aesthetic mastery. In demanding unity of effect and affect, Poe’s aesthetic represents an innovation in prose fiction and at the same time a narrowing of its vocal variability or plurality.

Benjamin counts Poe among his valued storytellers. This seems wrong to me on Benjamin’s own terms. The storyteller’s “gift is the ability to relate his life; his distinction, to be able to tell his entire life. The storyteller: he is the man who could let the wick of his life be consumed completely by the gentle flame of his story” (108-09). Poe? Benjamin is, I think, lured into feeding his nostalgia for storytelling with Poe’s stories because their first-person narrators and their brevity seem to fit his notion of the immediacy between teller and listeners. The novel is the death of storytelling in Benjamin’s view because it separates tellers and listeners from one another through the mediation of print: “The storyteller takes what he tells from experience-his own or that reported by others. And he in turn makes it the experience of those who are listening to his tale. The novelist has isolated himself. The birthplace of the novel is the solitary individual, who is no longer able to express himself by giving examples of his most important concerns, is himself uncounseled, and cannot counsel others” (87). But the mediation of print affects the inner form of Poe’s stories as much as the novels Benjamin has in mind. Moreover, the stylization through which Poe writes the storyteller precisely isolates the storyteller in an experience that is not communicated and shared so much as simulated and mastered in the aesthetic contemplation of reading. This stylization henceforth became available to modern novelists as a technique of aesthetic distance.

As a matter of literary history, Poe is part of the development of the novel not the preservation of storytelling, at least not in Benjamin’s sense. But this touches on the deeper problem in Benjamin’s historical claims. Bakhtin is truer to the history of the novel in seeing it as a continual appropriation of other social discourses, including the whole array of storytelling modes. The mediations of print and novelistic writing do not kill storytelling but appropriate and transform it in keeping with the exigencies of the modern forms and institutions of publicness. The nature of such novelistic appropriations and transformations is not uniform. Benjamin’s theoretical error is in supplying a single interpretation of the difference between storytelling and novel, setting a categorical divide between them.

His historical error is in supposing that the rise of the novel extinguishes storytelling from modem culture and social life. The ongoing development of the novel belies the declaration of the death of storytelling, just as it belies the many declarations of the death of the novel itself. The novel form continues to replenish itself and transform its own modes of cultural commentary by drawing on living-or disappearing-practices of storytelling in the construction of novelistic narratives. The sheer diversity of the appropriations suggests the complexity of the contemporary novel’s interaction with oral culture and the lifeworlds within which it flourishes. Maxine Hong Kingston draws on the “talk-stories” of immigrant Chinese life in America in constructing the innovative narrations of The Woman Warrior. John Edgar Wideman’s Hornewood trilogy elaborates its narratives and voicings out of the vernacular discourse of the black ghetto he writes about. Several strands of magical realism reinscribe the storytelling styles and norms of preliterate cultures in order to represent a profoundly modern experience of historical memory and change. The Indonesian novelist Pramoedya Ananta Toer “drafted” The Buru Quartet while a political prisoner by entertaining his fellow inmates with stories and then writing these tales down to make a complex historical Bildungsroman spanning several decades of Indonesian history; the very composition of his novel joins the storytelling context of prison life and the publicness of print culture, in defiance of the deformations of the public sphere by political repression and censorship.


Benjamin pegs The Sentimental Education as the supreme example of the novel in its utter negation of storytelling, citing and commenting on the closing passage in which Frederic and Deslauriers, middle-aged and shorn of all their aspirations and dreams, recall the boyhood episode in which they sneaked into the local bordello, presented the patronne with flowers, and then immediately fled:

“‘That may have been,’ said Frederic when they had finished, ‘the finest thing in our lives.’ ‘Yes, you may be right,’ said Deslauriers, ‘that was perhaps the finest thing in our lives.”‘

With such an insight the novel reaches an end which is more proper to it, in a stricter sense, than to any story. Actually there is no storyfor which the question as to how it continued would not be legitimate. The novelist, on the other hand, cannot hope to take the smallest step beyond that limit at which he invites the reader to a divinatory revelation of the meaning oflife by writing “Finis.” (100)

Benjamin pits the novelist’s “meaning of life” against the storyteller’s “moral of the story”; the latter’s wisdom stands off against the former’s perplexity: the novel’s “quest for [‘the meaning of life’] is no more than the initial expression of perplexity with which its reader sees himself living this written life.” In Flaubert, “the meaning which the bourgeois age found in its behavior at the beginning of its decline has settled like sediment in the cup of life” (99). But Flaubert’s “Finis” is even more pointedly ironic, more novelistic, than Benjamin claims. The finest moment which Frederic and Deslauriers here recall takes place three years before the beginning of the novel! It is outside the quest to which Flaubert gives form.

The irony is directed at Frederic insofar as he cannot in the end grasp any value in his life except outside the “adventure of interiority” (Lukacs) in which his longings and actions have taken place. At the same time, however, the irony turns back on the writer’s own ironic subjectivity, rendering him “as observer and creator … a free object of free irony,” for the very form he has given the quest threatens to collapse from its inability to bestow a meaning. Flaubertian irony lies in this vibration between Frederic’s empty gesture of endowing his life with meaning and the writer’s inability to extract a meaning from his own full rendering of that life. In reducing this permanent oscillation and uncertainty to some generalized life perplexity endemic to the novel as a bourgeois form, Benjamin misses the deeper validity of the novel as a form of social and cultural criticism, in Flaubert’s case a criticism of the bourgeois lifeworld itself.

Benjamin misses the critical force of the novel because he takes its commodity form in the marketplace and its exfoliation of modern. individuality as mere symptoms of capitalism. According to his history of forms, the novel is wedged between storytelling and the cinema, his nostalgia for storytelling counterbalanced by his utopian expectations for film. The ongoing interaction of storytelling, novel, and mechanical reproduction has proved more complex, more nonsynchronous, than predicted by Benjamin, who matched these different forms of cultural production to precapitalist, capitalist, and postcapitalist lifeworlds respectively. Leaving aside that the survival of capitalism incorporated. the film form as a major component of consumer culture (a historical development long recognized as qualifying the hopes Benjamin so boldly expressed in “The Work of Art in the Age of Mechanical Reproduction”), the novel form itself has drawn, in its spiraling fashion, more fully on the entire tradition of the novel than did its nineteenth-century antecedents and by the same token undergone unprecedented transformations in response to the mass media.


To illustrate my point, I want to take up an unlikely example to counter Benjamin’s sense of the novel form, namely, Norman Mailer’s Executioner’s Song. This “true-crime,” documentary novel explodes the categories by which Benjamin divides the novel from storytelling and from mechanical reproduction. In the process, it also achieves one of the most startling innovations in novelistic voice in contemporary literature.

The making of the book was no ordinary work of journalism. The journalist entrepreneur Lawrence Schiller-whose most recent contribution to American culture was peddling exclusive photos of O.J. Simpson’s arrival home after his acquittal-saw an opportunity for magazine articles and a film or book in the story of Gary Gilmore as soon as Gilmore began opposing any appeal or stay of his death sentence for two murders in Utah in 1976. He would become the first person executed in the United States since the Supreme Court lifted its longstanding ban on capital punishment. Schiller negotiated exclusive rights to the stories of virtually everyone involved in the case, from Gilmore and his girlfriend Nicole Barrett and their families and friends to the widows of the two victims. He set about taping scores of interviews and exercised considerable control over the media’s access to personal testimonies, private letters, and interviews. He created the oral archive of the Gilmore case, and it wasn’t until a month after Gilmore’s death that he made the deal for Mailer to write the book. Mailer came to the project in the role of a hack.

Mailer’s achievement as a writer in Executioner’s Song begins with his mastery of the sheer bulk of material, some 15,000 pages of interview transcripts, in addition to news stories, court records, and Gilmore’s autopsy. From these raw materials, Mailer constructs a narrative that turns around two profoundly troubling enigmas: the mentality of a sociopath whose own suffering and sensitivity, especially in his love for Nicole, can never be squared with his capacity to kill two people without motive or passion or remorse; and the mentality of a society that gleefully regresses to the rationalized irrationality of capital punishment, an irrationality made all the more glaring by Gilmore’s own insistence that he die.

The principles of construction at Mailer’s disposal are thoroughly novelistic. He plots a multi-character story, moving apace through the nine months between Gilmore’s release from an Illinois penitentiary and his execution in Utah, while braiding into that narrative the history of several of the characters. He holds the narrative together with the love story of Nicole and Gary, a story replete with menace, alcohol, sexual freedom and sexual dysfunction, tender letters, belief in reincarnation, and an abortive suicide pact.

What makes the novel, however, is its voice. Among Flaubert’s indelible contributions to the novel form is free indirect style, with its unlimited flexibility in evoking the subjectivity, the interiority and inner speech, of a character within the objectifying trajectory of third-person narration. The whole of Executioner’s Song is written in free indirect style. The underlying discourse of the characters comes from the 15,000 pages of transcript. Section by section, sometimes paragraph by paragraph, the source of the story switches from one character to another. Mailer’s prose is inflected with their perceptions and idioms. Written with lightning speed to meet the demands of a book market that might quickly tire of the Gilmore case, the novel’s scope and texture are made possible by the interviews. The writing depends on storytelling and on the mechanical reproduction by which those stories were recorded. The everyday storytelling of the small-town, working-class people who knew Gilmore during his brief time in Utah is the mainspring of the narration. Mailer brings to his retranscriptions the flattening tone and deadpan concision that give the novel’s voice its relative consistency across the multiple voices, in keeping with the norms of both journalism and Flaubertian free indirect style.

There is one storyteller who is indispensable to the narrative as a whole and whose voice resonates most centrally throughout, Nicole Barrett. Her life story, that of a free spirit in love with a succession of bad men, and the honesty with which she details the romance and hard facts of her relationship with Gary create a language that Norman Mailer couldn’t possibly invent, though he can write it:

Then she thought of the night up in the hills behind the nuthouse when she wondered if he was a magnet to evil spirits. Maybe he had to act that nasty to keep things 08 The idea didn’t cheer her. He could get meaner and meaner if that was the truth.

Around midnight, Nicole was feeling awfully cooped up with Gary.

She found herself thinking of Barrett. It kept working away in her. There had also been a letter from Kip that afternoon but she kept thinking about Barrett and Rosebeth.

She hadn’t even wanted to open Kip’s letter, and when she did, he wrote that he wanted her to come back. The letter left her feeling crowded. It was like the past was coming back. Hampton, of all people, was going around with her sister April. Everybody, Nicole decided, was fucking with her head.

All the while she was having these thoughts, Gary had been sitting at her feet. Now he had to pick this moment to look up with all the light of love shining in his eyes. “Baby,” he said, “I really love you all the way and forever.” She looked back. “Yeah,” she said, “and so do seven other motherfuckers.”

Gary hit her. It was the first time, and he hit her hard. She didn’t feel the pain so much as the shock and then the disappointment. It always ended the same way. They hit you when they felt like it. (158)

Who can deny this storyteller her experience and wisdom? Contrary to Benjamin’s diagnosis, the novel has proved capable not only of maintaining its vitality in the age of mechanical reproduction but also of preserving the very possibility of communicating experience that he nostalgically associated with storytelling.


Throughout this essay, I have mixed metaphors in my use of the terms voice and writing. I have argued that novelistic narration should be approached as an act of writing, but I have also retained the ambiguous concept of voice. Though the language of criticism is necessarily metaphorical, not all metaphors are created equal when it comes to their use and analytic power. Let me conclude then with a reflection on these two terms.

“Writing” is on the face of it a literal rather than metaphorical term. And indeed I hold the view, which I associate with the work of Bakhtin and Raymond Williams, that “literature,” broadly defined, is the social practice of writing and therefore inseparable from the social history of literacy. Nevertheless, poststructuralism threw a wrench into every purely empirical sense of “writing,” beginning with Jacques Derrida’s huge claim in Of Grammatology that Western philosophy conceptualizes speech and writing as opposites and then freights the concept of writing with whatever features of language are deemed errant and recalcitrant to the reigning metaphysical idea of the nature of language. De Man’s Allegories of Reading and Derrida’s own work as a literary critic, especially in the essays on Plato, Mallarme, and Philippe Sollers in La dissemination, revolutionized literary studies by showing that metaphors of writing are so integral to every practice of writing that it is impossible to say what writing is-as artistic activity, social practice, or vocation-without entering the metaphorical or figurative labyrinth of the written text.

When Rushdie writes the story of his fictional Sufiya Zinobia from a news item; when Wolf writes her autobiography by turning the “I” inside-out into third-person novel and second-person diary and memoir; when Morrison writes the story of Joe and Violet up to a denouement she does not foresee or fully comprehend; when Mailer writes hundreds of tape-recorded interviews about actual events into a unified novel in free indirect style-the critic cannot say what “to write” means except by struggling to grasp writing at once as social practice and metaphor-laden textuality.

“Voice” is overtly metaphorical and has acquired connotations ranging from onto-theological presence to the linguistic technicalities of grammatical mood. Yet, it has distinctive advantages over other widely used terms in novel criticism. Unlike “presiding consciousness,” it does not presuppose what shape the subjectivity of writing and reading actually takes, or ought to take, in novels. Unlike “point of view” or “perspective,” it does not import a visual metaphor into the account of a phenomenon of language. Narratology’s recourse to the visual metaphor is somewhat surprising. Gerard Genette’s goal in Narrative Discourse is to describe every aspect of the art of narration in rigorously linguistic and rhetorical terms, but when faced with novelistic prose’s capacity to create a voice that can simultaneously convey, usually tacitly, someone else’s attitude, responses, perceptions, vocabulary, or manner of speaking, he turns to a distinction between the narrative’s “voice” and the one who sees or perceives. This distinction gets worked up into a full-blown scheme in Mieke Bal’s theory of “focalization,” which she intends to clarify the “distinction between, on the one hand, the vision through which the events are presented and, on the other, the identity of the voice that is verbalizing that vision. To put it more simply: … between those who see and those who speak” (100-01). But is it really a clarification? In keeping with the boxes-within-boxes approach to narration, “focalization” theory uncritically recycles Cartesian res mentis and res extensa as though every layer-or box-of narration establishes a homogeneous relation between a perceiving subject and a world available to perception. Moreover, as I argued earlier, narrative theory thus grounds its critical categories in the created world of the novel (the world of the “story” in narratological terms) at the expense of the world of the “discourse.” Novel theory, by contrast, starts from this latter world, the actual worldliness of the novelist’s creative subjectivity and practice and of the novel’s cultural origins and uses.

It is significant’ that narrative theory’s commitment to linguistic and rhetorical rigor breaks down right where the problem of pinpointing the subject of novelistic narration encounters a multiplication of subjects (writer or implied author? implied author or narrator? narrator or character? voice or “focalizor”?). The palpable but indistinct intersubjectivity so essential to the stylistic complexity of novels haunts narrative theory. It cannot penetrate the shadowy world of narrators and characters because it eschews from the outset the concrete subject of novelistic writing, that is, the writer in all his or her overdetermined empirical social- linguistic existence.

Narrative theory draws on structuralist linguistics to outline the relation of the subject of the enunciation (speech event) to the e’nond (narrated event), but misses the most pertinent implications of the idea that the subject of the enunciation is referred to by means of the so-called shifters, principally, the pronoun I. In Emile Benveniste’s formulation, I is a signifier in the inonce’ which refers to the subject of the enunciation; it is completely devoid of semantic content, since it refers to this speaker only because it refers to whoever is speaking. The linguistic marker of identity is anonymous. What follows from this? Structuralist narrative theory and narratology conclude that the subject is therefore but an effect of discourse or the outer limit of the narrative boxes and therefore moot. I think that Benveniste implies, rather, that the subject of the enunciation is at once the referent of an empty signifier and the concrete subject who produces the discourse. Therefore, we only glimpse this subject’s presence fleetingly in the movement of his or her enunciation as a whole.

Here Bakhtin’s insistence that “the real unit of speech communication” is the “utterance” takes on its full import for literary studies (Speech 71). An utterance-or “speech event” (enunciation) in structuralist terminology-is demarcated in the moment it tacitly calls for a response; its length can vary “from the single-word rejoinder to a large novel” (81-82). It is in that sense that the subjectivity of the writer can only be glimpsed in the reader’s responsiveness to the movement of the utterance as a whole. The aesthetics of the novel turns on this worldly relation between concrete subjects. The critical categories of novel theory and narrative theory traverse “discourse” and “story” in opposite directions. In the analysis of first-person narratives, it is the difference between “a story recounted by a narrator contemplated by an implied author” and “the author writing the narrator telling a story.” The worldliness of writers and readers, which remains the bracketed-out limit of narrative theory, is the starting point of novel theory. “Of course,” writes Bakhtin, “these real people, the authors and the listeners or readers, may be … located in differing time-spaces, sometimes separated from each other by centuries and by great spatial distances, but nevertheless they are all located in a real, unitary and as yet incomplete historical world set off by a sharp and categorical boundary from the represented world in the text” (Dialogic 253).

At stake is how criticism understands literary creativity. I follow Bakhtin’s assertion that the “real, unitary and as yet incomplete historical world” is “the world that creates the text, for all its aspects-the reality reflected in the text, the authors creating the text, the performers of the text (if they exist) and finally the listeners or readers who re-create and in so doing renew the text-participate equally in the creation of the represented world of the text. Out of the actual chronotypes of our world (which serve as the source of representation) emerge the reflected and created chronotypes of the world represented in the work (in the text)” (253). Because he does not separate language from its worldliness, Bakhtin does not put the reality that creates the text or the reality to which it refers outside language. Reality is not extratextual. As I argue in “On Innovation,” the realist imperative continues to animate the contemporary novel’s formal and stylistic innovations. Unlike narrative theory, which often seeks to reconnect formal structures and social categories after having strictly separated them, novel theory is not constrained to smuggle social reality back into textual analysis.

Moreover, because novel theory starts from the worldly intersubjectivity of writing and reading, it eliminates the need for dubious categories like “vision” or “focalization” to fit intersubjectivity into novelistic prose, whether the shadowy intersubjectivity of free-indirect third-person narration or the relatively delimited intersubjectivity of unreliable first-person narrations. For Bakhtin, language is in essence intersubjective, lying “on the borderline between oneself and the other. The word in language is half someone else’s. It becomes ‘one’s own’ only when the speaker populates it with his own intention, his own accent, when he appropriates the word, adapting it to his own semantic and expressive intention” (293). The novelist’s stylistic achievement of “double voicing,” however masterful, derives from the existential condition of every speaking being: no one ever truly originates or masters speech.

In such a conception, novel theory deploys the metaphorical concept of “voice” without falling prey to the metaphysics of presence that deconstruction decries or to the simulated Cartesianism of narratology’s narrators and narratees, speakers and “focalizors.” At the same time, it does not repudiate the common use of the metaphor, as when we say that an author has created a “distinctive voice” or that a work of prose “has a voice.” Such statements may seem naively intuitive, but they in fact point the way to questions that criticism has still barely explored. For example, how does a writer’s stylistic realization of a “voice” shape the interplay of identity and anonymity intrinsic to language? And how does our aesthetic response to novelistic “voice” relate to the everyday experience of recognizing someone’s voice, even to the experiences of infancy that lead Jacques Lacan to say that the voice is a primordial libidinal object? “We always arrive, in the final analysis,” Bakhtin was not afraid to say, “at the human voice, which is to say we come up against the human being” (Dialogic 252-53).