Journal of Narrative Theory

This paper sketches out some methodological coordinates for investigating the formal category of narrative voice in a broader discursive context. It seeks to reformulate the classic model of narrative communication in order to redress the imbalance of current narratological scholarship, which focuses on theorizing the role of real readers without due attention to real authors. I have developed this approach to investigate and explain a specific problem: how to account for the increased prominence of omniscient narration in literary fiction over the last two decades. Does contemporary omniscience differ from the classic omniscience of eighteenth- and nineteenth-century fiction, and if so, what does that difference say about the cultural status of the novel in current public discourse?¹

I’ll begin by illustrating this problem through a brief discussion of narrative voice in William Makepeace Thackeray’s 1848 Vanity Fair and Martin Amis’ 1995 The Information. At one point in Thackeray’s novel, the narrator pauses to address readers as follows:

If, a few pages back, the present writer claimed the privilege of peeping into Miss Amelia Sedley’s bedroom, and understanding with the omniscience of the novelist all the gentle pains and passions which were tossing upon that innocent [End Page 91] pillow, why should he not declare himself to be Rebecca’s confidant too, master of her secrets, and seal-keeper of that young woman’s conscience?

(171).

By contrast, Amis’ narrator laments, “And I made the signs—the M, the A—with my strange and twisted fingers, thinking: how can I ever play the omniscient, the all-knowing, when I don’t know anything?” (63). Both passages exemplify intrusive omniscient narration in that the narrators reflect on their own authority as storytellers, presenting themselves not just as narrators, but as novelists and authors of the books that we are reading. Each of these examples makes specific reference to the function of literary omniscience as a form of knowledge, about which there is an ongoing debate sparked by Jonathan Culler’s 2004 article in Narrative, entitled “Omniscience.” In my own contribution to that debate, I discussed how its parameters have been largely epistemological and theological: limited to asking how and how much an omniscient narrator “knows” about the fictional world, and what sort of narratorial figure or entity can be considered omniscient (Dawson 143–61). In a sense, Thackeray’s and Amis’ narrators are grappling self-reflexively with these same questions. What interests me in this essay, though, is not the questions themselves, but why the narrators foreground them.

If we conduct a classic taxonomic study of these two novels, we will see that both narrators display all the knowledge of their respective fictional worlds characteristic of omniscient narration, including variable (or zero) focalization, access to consciousness, and spatio-temporal freedom. In terms of Gérard Genette’s category of mood, or the various means by which information about the story world is regulated, the novels differ little, although The Information is less panoramically ambitious, focalizing mainly through the protagonist Richard. In terms of Genette’s category of voice, or the narrating instance, the novels also differ little: the person, time, and level of the narrating are all the same. So if we tick off the list of their formal properties, we can classify synchronically these novels as omniscient. Yet surely there is a palpable difference between the performative stances that these two narrators adopt. In the Thackeray passage, the novelist confidently and playfully asserts the privilege of omniscient knowledge, whereas in the Amis passage, the narrator manifests anxiety about that omniscient authority. In fact, Amis’ narrator grapples not with a failure [End Page 92] of diegetic knowledge, but with a failure of novelistic insight resulting from his own limitations as a person. He reflects scenically on his own experience in order to ask whether he can satisfy his role as an observer of human nature.

If there is a formal difference between these two examples of omniscient narration, it rests in Genette’s last and least developed element of voice: the function of the narrator.² Both of these novels enact what for me is the key feature of literary omniscience: the performance of narrative authority through intrusive narratorial commentary, which “personalizes” the narrator as an extra-diegetic character. What distinguishes omniscient narration from other heterodiegetic narratives is the clear distinction that it makes between narratorial and characterial perspective in the ideological, not spatio-temporal, sense, by which I mean the consistent rhetorical strategy of that narration, manifested most overtly in intrusive commentary, not simply the potential irony emerging out of free indirect discourse.

The differing functions of Thackeray’s and Amis’ narrators stem from the distinct ways in which they establish narrative authority as omniscient storytellers, a divergence that can only be explained by reference to the historical shift in the cultural status of the novel from the mid-nineteenth to late-twentieth century. In his 1966 study The Form of Victorian Fiction, J. Hillis Miller argues that the Victorian convention of omniscient narration “is so crucial to nineteenth-century English fiction, so inclusive in its implications, that it may be called the determining principle of its form” (63). More recently in Bad Form, Kent Puckett claims that the nineteenth century “saw both the European novel and an omniscient narration whose voice was the voice of that novel’s cultural authority come into their own” (6). If omniscient narration is so closely identified with the status and form of the nineteenth century novel, what kind of authority do contemporary novels with this convention invoke? To answer this question, we need to investigate the rhetorical strategies employed by authors in a period when the weight of critical opinion and literary fashion suggests omniscience is an outmoded narrative voice that cannot speak with the same authority to a contemporary readership. No longer simply a convention of storytelling from which to choose amongst others available to writers, this narrative voice carries the weight of its association with the supposed high period of the genre of [End Page 93] the novel itself: an age before the competing claims of new media forms (from radio, to cinema, to digital technology) when the public sphere still sought guidance in ethical conduct from literature and novelists could assume, in the rhetoric of narration, a shared set of cultural values with their readers.

In the passage from Vanity Fair above, the narrator follows his rhetorical question about access to consciousness by providing an account of Becky’s thoughts and of the social context required for the written “history.” In describing Becky’s regret over turning down a marriage proposal that would have secured her a prosperous life and high social status, the narrator engages the narratee directly:

What good mother is there that would not commiserate a penniless spinster, who might have been my lady, and have shared four thousand a year? What well-bred young person is there in all Vanity Fair, who will not feel for a hardworking, ingenious, meritorious girl, who gets such an honourable, advantageous, provoking offer, just at the very moment when it is out of her power to accept it? I am sure our friend Becky’s disappointment deserves and will command every sympathy.

(171)

Thackeray’s narrator solidifies this (perhaps ironic) identification with Becky’s plight by reference to his own experience, when he claims: “I remember one night being in the Fair myself, at an evening party” (171). In this breach of the story/discourse division, the narrator relates a firsthand observation of how an impending marriage into a higher social rank can alter a person’s standing: “If the mere chance of becoming a baronet’s daughter can procure a lady such homage in the world, surely, surely we may respect the agonies of a young woman who has lost the opportunities of becoming a baronet’s wife” (172).

This section is a good example of Miller’s claim that the Victorian narrator is immanent rather than transcendent, possessed of an omniscience that moves within the community of the narrated story. Miller draws attention to the quality of Thackeray’s omniscient narrator that “identifies him as a perfect example of a spokesman for the general consciousness of the community”: [End Page 94]

This is his use of the editorial “we.” The novel is punctuated by direct addresses to the reader in which he is encouraged to think of himself as one of a vast number of other readers who share similar experiences of life and similar judgments of it. We are asked to identify ourselves with one another and with the narrator who speaks for us until by a kind of magical sympathy we lose our identities, are drawn into the group, and taken all together come to form a ubiquitous chorus of judgment.

(72)

Miller calls this quality a “rhetoric of assimilation,” which in “establishing the reader’s participation in a community mind surrounding the individual minds of the characters in the story gives the strength of a universal judgment” (72, 78). So this extra-diegetic appeal to a common reading public supplements the diegetic authority, the omniscient knowledge, of Thackeray’s narrator, established in the prefatory chapter as a puppeteer, “the Manager of the Performance.” This “community mind” is obviously a rhetorical construct rather than a sociological fact, and the judgments that it endorses, as the preface intimates, are those of a “man with a reflective turn of mind” in sympathy with the narrator’s own stance towards the characters. In her study of “gendered interventions” in Victorian fiction, Robyn Warhol describes Thackeray’s particular mode of direct address as a strategy of “distancing,” most common in novels written by men, in contrast to the direct address of “engaging” narrators, most common in novels written by women. For Warhol, “a distancing narrator discourages the actual reader from identifying with the narratee, while an engaging narrator encourages that identification” (31).

The editorial “we,” whether distancing or engaging, is largely absent from contemporary omniscient narratives, as it is from Amis’ The Information. Amis’ narrator is as intrusive as Thackeray’s but does not, and cannot, invoke a community mind. He thus requires recourse to a different means of character evaluation. The protagonist Richard Tull is a failed writer, his dedication to avant-garde experimentalism heightened by the obscene popular success of his friend with a work of middlebrow fiction: “Essentially Richard was marooned modernist,” the narrator tells us, “Modernism was a brief divagation into difficulty; but Richard was still out there, in difficulty. He didn’t want to please the readers” (170). [End Page 95] Richard’s struggle as a writer, which provides the narrative momentum of the novel, is in fact a struggle over the concept of the universal. In an argument about whether Richard will finish his novel, and whether it will end up being published and earning money, his wife Gina says, “I don’t know if you still really believe in it. Your novels. Because you never . . . Because what you . . . Ah I’m sorry, Richard. I’m so sorry” (87). A brief paragraph of narration follows this line of dialogue, completing Gina’s unfinished sentence: “Because you never found an audience—you never found the universal or anything like it. Because what you come up with in there, in your study, is of no general interest. End of story. Yes, this is the end of your story” (87–88). This passage could be the narrator’s rendering, via omniscient knowledge, of Gina’s unvoiced thoughts, or it could be the rendering of what Richard thinks are her unvoiced thoughts, doubling as an internal dialogue with his own self-doubt. It could also be the narrator’s address to his character, for the narrator struggles throughout the book with the ‘universal’ authority of his own omniscience, echoing Richard’s anxiety both ironically and agonistically.

Richard’s diminishing sexual potency parallels his difficulties in finishing his book (at one point, he uses his anxiety over “the death of the novel” as an excuse to his wife for his poor sexual performance), and much of the book deals with his mid-life crisis. In one scene, Richard experiences a spontaneous erection while his young son moves about innocently on his lap:

This used to cause him disquiet, and struck him as something he had better shut up about. But, again, he was enough of an artist to have faith in the universality of his own responses. He asked around among the dads and found that it was so. It was general—universal. It still stuck him as essentially perverse. When you thought of all the other occasions which cried out for hardons that never came. And here you not only didn’t need one. You didn’t even want one.

(195)

Whereas Thackeray’s narrator describes himself in the book as “an observer of human nature”—moving through Vanity Fair in person and drawing upon the novelistic convention of omniscient access to character [End Page 96] thought to supplement his moral commentary—Amis’ narrator makes more introspective observations (177). The effect is not to undermine his authority, but to ground the legitimacy of his observations in his own experience. The passage above, in which the narrator wonders how he can be omniscient when he doesn’t know anything, comes from a section of commentary beginning, “This whole thing is a crisis. This whole mess is a crisis of the middle years” (62). In what follows, it becomes obvious that this crisis refers not only to Richard’s life, but to the book itself—including perhaps its genesis, if we wish to read it autobiographically, yet more importantly its form. The next line reads, “Every father knows the loathed park and playground in the unmoving air of Sunday morning (every mother knows it Friday evening, Tuesday afternoon—every other time)” (62). The narrator’s own account of a time when a child approaches him in the playground and proceeds to spell out letters in sign language personalizes this “universal” comment about parenthood. The narrator believes that the child is deaf and dumb, and leans forward, attempting to decipher the letters, “suddenly braced for revelation, frowning, essaying, as if the boy could tell me something I really might need to know” (63). The boy’s announcement that he has spelled out his own name precipitates the narrator’s crisis of omniscience, inducing a sense that he lacks the knowledge of human nature, the “information,” necessary to write a novel. His strategy is not to build this sense of crisis into the structure of the novel in metafictional fashion, but to reassert his authority through a confessional identification with his character: “I wrote those words five years ago, when I was Richard’s age. Even then I knew that Richard didn’t look as bad as he thought he looked” (63). Amis’ narrator, then, confesses to readers rather than engaging them in a dialogue. He cannot assert the “universal” by general observation or the assumption of collective agreement; he can only offer it provisionally, through individual introspection, which is what allows him to claim: “Intimations of monstrousness are common, are perhaps universal, in middle age” (64).

Clearly, Amis’ narrator is gendered, as his strategy is to identify with Richard’s perspective. This strategy foregrounds the relativity of the narratorial commentary, which is replete with stereotypical statements about gender differences: “She was a woman. She knew so much more about tears than he did” (9). One can see at work here a disavowal of the role [End Page 97] that gender politics has played in rendering the concept of the ‘universal’ untenable. Vera Nunning locates The Information within a trend of contemporary fiction that she describes as a “merging of realism and experiment” (249). According to Nunning, Amis “parodies nearly all the characteristics of nineteenth century authorial narration and refuses to conform to the dogmas of political correctness” (249). I would suggest that The Information is less a parody than an agonistic encounter with those characteristics, and that its refusal of “political correctness” is an element of this agon. For instance, after describing the beauty of Gina (focalized through Richard), the narrator offers this version of the editorial “we” in a search for universality:

We are agreed—come on: we are agreed—about beauty in the flesh. Consensus is possible here. And in the mathematics of the universe, beauty helps tell us whether things are false or true. We can quickly agree about beauty, in the heavens and in the flesh. But not everywhere. Not, for instance, on the page.

(15, emphasis original)

The difference between the performance of narrative voice in Vanity Fair and The Information, novels emblematic of the nineteenth and late twentieth century respectively, requires a diachronic account of shifting modes of omniscient authority. This account is not so much of the historical decline and revival of a narrative convention (hence mapping onto a standard evolutionary model of the progression of the novel from authorial to figural narration), but of the historical mutability of that convention. And if this mutability resides largely in the narrative function of commentary, then it requires a synchronic investigation of the discursive relation between narrative voice and the contemporaneous nonliterary discourses in that particular period.

Given this context, and my claim for locating omniscient authority in the function of the narrator, we can see how the narrating instance invokes a historically specific figure of the author. Theories of omniscience tend to revolve around information and privilege, the domain of focalization, when they should be firmly located in the domain of narrative voice. There is no space here to rehearse the myriad and complicated debates over the relationship between voice and focalization, but I will make the claim that [End Page 98] focalization or perspective in the broadest sense should be assimilated into the category of voice and approached as a rhetorical strategy of the narrator.³ This analytic hierarchy is evident in Manfred Jahn’s frame model of the narrative situation (adapted from Mieke Bal): the narrator tells the narratee that the focalizer sees that the focalized does. Jahn’s description of “an authorial narrator who does not participate in the action, who is normally omniscient, omnipresent, and reliable, and who is responsible for exposition, temporal organization, choice of narrative mode, rhetoric and style (usually well-spoken)” clarifies the operation of this hierarchy in omniscient narration (445, emphasis added).

In the introduction to New Perspectives on Narrative Perspective, Willie Van Peer and Seymour Chatman argue for a narratology that moves beyond product-analysis to process-analysis, which “changes the basic research model from the question, What is the perspective in this story? to How is perspective in this story brought about?” (7, emphasis original) Contemporary narratology approaches the question largely by focusing on how readers process perspective. Yet if we hearken back to pre-narratological concerns with novelistic method, and ask how authors bring about perspective in a story, we can then ask why: for what broader cultural purpose do authors construct narrators who employ different types of focalization as part of a rhetorical assertion of narrative authority?

Classical narratology, as we know, posited as its object of study a series of agents immanent to a work of narrative fiction while leaving aside the agents who bracket that formulation: the author and the reader. Chatman’s diagram of this model remains the standard:

Click for larger view
View full resolution

Shlomith Rimmon-Kenan, in Narrative Fiction, argues that: “Of the six participants enumerated in this diagram two are left outside the narrative transaction proper: the real author and his equally real counterpart, the real reader” (87). She goes on to dismiss Chatman’s claim for the importance of the implied author and reader (and his disregard for the narrator and narratee): [End Page 99]

Only four of Chatman’s six participants are thus relevant to my conception of narration: the real author, the real reader, the narrator, the narratee. Furthermore, as I have suggested in the introduction, the empirical process of communication between author and reader is less relevant to the poetics of narrative fiction than its counterpart in the text. This chapter will therefore deal with two participants only: the fictional narrator and the fictional narratee.

(90)

In a recent book chapter, entitled “Why Don’t Our Terms Stay Put? The Narrative Communication Diagram Scrutinized and Historicized,” Harry E. Shaw notes that “users of the diagram bring to it two different implicit models of the communication situation” and that “the terms the diagram seeks to describe necessarily become hazier as we move from left to right” (299). If the communication diagram evokes “the image of someone telling a story to someone else,” Shaw claims, the first model emphasizes the flow of information between the various communicative agents, while the second emphasizes the “effects and purposes the teller wishes to achieve” (300).

Shaw’s distinction between information and rhetoric can also be seen as a distinction in analytical emphasis between focalization and voice. In the communication model, narrators and narratees belong to the discourse, whereas the focalizer and the focalized belong to the story. Information theorists would see the narration as the medium through which readers access or construct a focalized story. Rhetoric theorists would see the focalized story as a vehicle for establishing a narrative effect. Given that story is only an effect of discourse, we must surely privilege the latter.⁴ Despite Shaw’s emphasis on the importance of the narrator, postclassical narratology has engaged primarily with the hazier right-hand side of the diagram. The story of postclassical narratology may be seen as an attempt to take up the challenge of theorizing the bracketing agent on the far right of the diagram, what is variously called the real reader, the actual reader, the empirical reader, and the flesh-and-blood reader. Two prominent approaches to this challenge are the rhetorical-ethical and the cognitive. One focuses on the ethical judgments that readers make in response to the rhetorical techniques employed in a narrative; the other investigates how readers process and make sense of narrative elements to construct mental storyworlds. [End Page 100] Both betray a tension between academic and general readers even as they try to elide the distinction.

In the rhetorical approach, the critic stands in as a test case for the flesh-and-blood reader, one capable of entertaining a range of cognitive, emotional, aesthetic, and ethical responses, while still asserting a final critical judgment. For instance, in applying his model of rhetorical poetics to Ian McEwan’s Atonement, James Phelan writes, “I believe that flesh-and-blood readers who respond in these ways are missing some of the intricacies of McEwan’s communication, but I also believe that in McEwan’s strategies they have good reasons for their responses” (“Delayed” 131). This emphasis on readers begs the question: how do we know the ways in which readers respond to a narrative? This question has animated criticisms leveled at the related field of reader-response theory, which have pointed out that the reader under study is typically an idealized version of the (androcentric) critical self.⁵ In a review of Phelan’s Living to Tell About It, Michael Eskin claims that Phelan cannot generalize his claims about the range of readers’ responses to narratives into a theory of narrative unless he incorporates an empirical approach into his critical practice.⁶

The cognitive approach in narrative theory attempts to redress this methodological shortcoming by overtly drawing its theoretical authority from empirical research in cognitive science. In doing so, it collapses reader responses to narrative into universal shared mental processes, while still maintaining a scholarly distance to apprehend these processes in operation and articulate their relation to the narrative text. Jahn, in his cognitive approach to third-person narration, warns against distinguishing between professional and general readers: “Ultimately, . . . juxtaposing of a sophisticated narratologist’s reading and a general reader’s reading highlights in a rather unflattering way the detrimental effect of mainstream narratology’s failure to account for what should be one of its prime considerations, the cognitive mechanics of reading” (464). Jahn goes on to argue that narratological readers not only share the same cognitive mechanics as general readers, but must embrace this shared process of reading in order to generate more sophisticated textual analysis:

Despite the fact that recourse to readers, readers’ intuitions, and reading plays an important part in narratological argument, [End Page 101] the contribution of mainstream narratology is preoccupied with bottom-up analyses, often assuming determinacies in violation of the Proteus Principle and indeterminacies in the presence of established cognitive preferences.

(465)

The cognitive study of narrative most reliant on empirical research is the psychonarratological approach of Marisa Bortolussi and Peter Dixon founded on the view that “[h]ow readers process narrative is essentially an empirical question that can only be answered by systematic observation of actual readers reading actual texts” (13). In distinguishing between “textual features (i.e., objective and identifiable characteristics of the text) and reader constructions (i.e., subjective and variable mental processes),” Bortolussi and Dixon make a series of hypotheses about how readers will respond to a specific textual feature and then test those hypotheses by conducting “textual experiments, in which particular features of a text are identified and manipulated by the researcher” (37, 51).

This method provides a bridge between narrative theory concerned with “real” readers and broader empirical studies of readers and the reading process. In a 2006 article entitled “Empirical Approaches to Studying Literary Readers: The State of the Discipline,” which locates the work of Bortolussi and Dixon in this field, David Miall claims that the “serious commitment to the examination of reading and the testing of hypotheses about reading with real readers . . . differentiates it clearly from the reader-response studies of the last thirty years, from Fish to Wolfgang Iser” (307). An important aspect of this differentiation, for Miall, is that empirical approaches sharply distinguish real readers from “professional” readers who produce published interpretations of literary texts. Real readers are “non-professionals” whose “ordinary literary reading” may be unconcerned with interpretation (294). Significantly, while acknowledging the importance of a “reader’s particular identity and cultural situation,” Miall argues that the processes which precede and support any act of interpretation “themselves are constituted by the cognitive and affective equipment that we possess in common with our reading ancestors” (293). According to Miall, “empirical studies of readers and reading” have the potential to “provide new landmarks for a more socially responsible and ecologically valid form of scholarship” (307). The problem with this sort of claim, I [End Page 102] would argue, is that the empirical approach remains open to the charge that it does not study real readers so much as lab-rat readers. Miall points out that “[o]ften experimental methods involve laboratory conditions in which acts of reading can be controlled and monitored” and that “[t]ypically, the readers studied will be drawn from the student population” (292). The category of the ‘real’ reader can thus be seen as a virtual construct of literary theory, which seeks to corroborate and universalize the professional theorist’s critical response to a text under the guise of testing how readers actually read.

The limitations of this approach are taken up in a 2009 special issue of Language and Literature devoted to an ethnographically oriented and thus, it is claimed, more ecologically valid study of reading. In their contribution to this issue, Joan Swann and Daniel Allington distinguish between two methods for studying real readers: the “experimental” and the “naturalistic.” The first, they suggest, involves “the artificial environment of a reading experiment,” generally taking students and testing “pre-specified and isolated aspects of reading” (248). The second, which they favor, involves investigating readers “in their usual environment, engaged in habitual reading behavior” (248). To provide a case study of “social reading,” Swann and Allington observed the interpretations and evaluations of literary texts that readers made in reading-group environments. In the analysis of their findings, they emphasize the role of interpersonal discussion and the “culturally and historically contingent” nature of reading contexts. While Swann and Allington draw a contrast between experimental and naturalistic studies “in terms of research design and focus,” these approaches share the same definition of real readers, which they characterize as “‘ordinary readers’—i.e. readers other than academic critics and professional reviewers” (260, 248). If we are seeking the “ecological” value of empirical studies of reading, however, it would seem unproductive to dismiss published “interpretations” by professionals in favor of ordinary acts of reading by nonprofessionals, especially given the influence of “unreal” professional critics and reviewers on the publication and reception of literature, and hence upon practices of social reading. A comprehensive account of the ecology of reading would distinguish between these types of readers, but also incorporate each of them into its analysis.

Ultimately, we must recognize that the reader is a methodological construct, one emerging out of the research questions that we pose. So how [End Page 103] might we theorize the real reader when accounting for the narrative authority of contemporary omniscience? The empirical approaches outlined above concern themselves with the cognitive and affective mechanics of reading and its social and interpersonal dynamics. My concern is with the public reception of literary works. I therefore intend to reconsider the narrative communication model by articulating an approach to narrative that acknowledges fictional narratives as public statements in a broader discursive formation and therefore as vital elements of public discourse. Such an approach does not proceed from a distinction between what is inside a narrative text and what lies outside it, but treats the narrative discourse of fictional texts alongside other nonfictional and nonliterary discourses in the public sphere. Here I am betraying the influence of Mikhail Bakhtin and, especially, Michel Foucault. In a sense, I am trying to negotiate a link between Bakhtin’s belief in authorial agency, in that person who orchestrates public discourses in the novel, and Foucault’s assertion that we must avoid seeing literature as a substitute or “general envelope for all other discourses” (“The Functions” 308). My aim is to investigate the ramifications of this discursive approach for a narratological theory of authorship, particularly one that takes into account authorial responsibility and narrative authority in relation to contemporary omniscient narration.

It is interesting to note that by the mid-twentieth century, when the modernist ideal of effacing the presence of the author became entrenched as an aesthetic principle in literary practice, it was buttressed by both the “intentional fallacy” of the New Criticism and the fundamental narratological distinction between author and narrator. Yet at the same time, the presence of the author in the public sphere became increasingly important to the marketing of fiction, as evidenced by the advent of the Paris Review interviews in the 1950s, establishing the genre of the author interview, and the emergence of writers’ festivals around the globe.⁷ These forums call upon authors to explain the genesis and motivations of their work and comment on the broader social issues with which their fiction engages, as if to supplement what cannot be made overt in the fiction. They enable authors to establish a public presence alongside their own non-fictional journalistic and essayistic writing.

Recognizing this literary ecology, then, requires approaching fictional narratives not as mediums for private or abstract communication between individual authors and readers, but as public sites of negotiation between a [End Page 104] range of subject positions. In adopting this approach, I wish to develop a model of narrative communication that reconceptualizes the various agents of this model as subject positions anchoring textual utterances in the public sphere. In a 2007 article, James Phelan discusses the unreliable narration of Lolita in order to “account for two especially notable groups of readers”: those who are seduced by Humbert’s narrative voice and those who are not (“Estranging” 223). What is implicit in Phelan’s rhetorical approach to readers’ responses is what I want to make explicit: the concrete evidence of actual readers’ public responses that can be situated alongside the narrative discourse of fiction. The textual forms of those responses range across three overlapping forums: the literary establishment, in the form of reviews and feature articles; academia, in the form of scholarly essays and monographs; and the general public, in the form of letters, blogs, online forums, and customer reviews. Together these constitute empirical textual evidence of the reader as a public reader, a figure that has the most material impact on the survival of a book. Such an approach is important for understanding that narrative authority is not purely immanent to a text, something that can be recuperated from a formalist study of narrative conventions, such as privilege or level. Culturally speaking, readers must grant narrative authority, and this attribution of authority does not occur solely through their cognitive processes as individual agents of textual perception. That authority is contingent upon the collective public textual response to the narrative in question.

The corollary of this approach is that to understand the modes of narrative authority specific to contemporary omniscient narration, we must investigate the rhetorical strategies that authors employ as public figures, not just those employed by narrators. Narratology has long eschewed consideration of authorship, except in the controversial guise of the implied author, originally proposed to unyoke the question of “intentionality” from its relation to authorial biography. The implied author is a way of providing an anthropomorphic centre for a narrative, even if there is an effaced narrator or no narrator, for it attributes implicit personal values and norms to the design of the narrative itself or, at least, acknowledges that readers construct an authorial persona out of the text. Theories of implied authorship have undergone a range of permutations since Wayne Booth’s original formulation and, in many cases, have moved so far from questions of authorial agency that some claim that each reader constructs a different [End Page 105] implied author to guide and affirm his or her reading. For instance, Wolf Schmid writes in his survey of debates over this concept, “it must be remembered that, like the readings of different recipients, the various interpretations of a single reader are each associated with a different implied author” (161). Ansgar Nunning even proposes that “a pederast would not find Humbert Humbert, the fictitious child molester and narrator of Nabokov’s Lolita, unreliable” (97). Such contentions leave us with the problem of weighing the relative significance of individual private acts of reading and the general public reception of a text.

My question here is whether readers “know” that the authorial persona is only that of an implied author when they infer it from a fictional text. That is, are they complicit in their own construction of an imagined entity, or do they infer what they think is the real author? If we wish to posit an implied author as a mediating entity between author and narrator, we need to define what we actually mean by a real author. We typically define the author as the historical figure who wrote the book and then spend time our debating the existence of narrators and implied authors.⁸ The author emerges as an aporia, granted both an existential solidity and an epistemological evanescence, disappearing from our knowledge in the act of reading. The problem is that we are dismissing a straw man concept of authorship: a figure with singular intentions and coherent values and norms.

In defense of the idea of an implied author, some claim that an author assumes a particular persona when writing a book, in the way that we assume a professional persona when writing a job application, and that we must therefore be careful to distinguish that ideal, or at least different, self from the real author (see Peter Rabinowitz). In promoting the continued ethical importance of the implied author, Booth writes, “In every corner of our lives, whenever we speak or write, we imply a version of our character that we know is quite different from many other selves that are exhibited in our flesh-and-blood world” (“Resurrection” 77). This proposition, for me, is a kind of endless deferral of the “real” to a zero point of an essentialized private self, one only ever accessible in the “flesh and blood.” Or it is the opposite: an evacuation of any sense of a knowable self in favor of a series of performative selves that we all construct for different social occasions, in which case an authorial persona cannot be any less “real” than any other self that writers adopt in their lives, or those of a real reader. Regardless, readers surely infer the figure of an author, a public [End Page 106] figure whom they hold responsible for the book that they are reading, rather than simply a private citizen whose personal values and norms underpin the narrative, and they construct that figure not only from the narrative text, but from extratextual elements.

The rationale for positing an implied author is that the narrative text mediates communication between authors and readers and hence allows no direct access to an author’s intentions. But, as I have pointed out, such access is available in the actual world of the public sphere through the author’s interviews and non-fictional writing, and readers show enormous interest in the voice of the author at writers’ festivals and readings. What Phelan calls the “recursive relationship” between authorial agency and reader response is therefore not just facilitated by textual phenomena, but by the author’s and readers’ extratextual statements circulating alongside the fictional text in the public sphere (Experiencing 4). If readers do construct an implied author, it is only to facilitate their response to the real author, which means that we need to attend to the crucial function of real authors not simply as producers of a narrative text, but as active participants in the process of reception.

A discursive approach begins with the assumption that a key challenge of contemporary narratology is how to negotiate methodological relations between formalist approaches to textual features and contextualist approaches to the contingencies of textual production and reception. Proceeding from an understanding of fiction as public discourse, how might we incorporate extratextual public statements of authors and readers in the narrative communication model to develop a theory of authorship? I aim to do so by refining the two major narratological approaches to fictional texts as published books active in the public sphere rather than as static formal artifacts: that proposed by Susan Lanser in The Narrative Act (1981) and that proposed by Genette in Paratexts (1987).

The authority of a published text, Lanser argues, is vested in what she calls its extrafictional voice, “the most direct textual counterpart for the historical author,” which “carries all the diegetic authority of its (publicly authorized) creator and has the ontological status of historical truth” (122, emphasis original). Now this formulation, as Lanser points out, is similar to the implied author; however, she locates it not in the narrative discourse, but in the extrafictional elements of the material book itself, from chapter divisions to authorial prefaces and publication details. Lanser is [End Page 107] content to focus on how readers respond to this extra fictional voice, but the very concept provides a methodological point of departure for theorizing authorship more broadly in relation to the narrative communication model. It enables us to approach the author not as a private citizen speaking to readers through the narrative text in order to convey personal values and norms, but as a public intellectual discursively engaging the reader by linking narrative and extrafictional voice. “The extrafictional voice,” Lanser argues, “is the most immediate vehicle available to the author, and although most novelistic communication does not take place on the extrafictional level, the extrafictional voice carries more than its quantitative proportion of impact” (128).

Lanser indicates that the extrafictional voice reconstructed from textual information within the book itself differs from “extratextual sources of information about the author or the book” (124, emphasis original). I would argue, however, that if this extrafictional voice frames the text and its narrative discourse, it also turns the text outwards to the broader public sphere and its range of extratextual sources. Here we find other vehicles of communication available to authors: public statements ranging from essays and manifestos to interviews and opinion articles, which together constitute a rhetorical strategy establishing their literary authority in public discourse. For Lanser, the “author” is “a textually encoded, historically authoritative voice kin to but not identical with the biographical person who wrote the text” (152). In which case, this “textually encoded” authorial voice must be constituted by both the extrafictional voice of a book and the author’s extratextual material. But it is also constituted by the narrative voice of the author’s various fictional works, for these generate the author’s status as a public figure. Here we see the value of Foucault’s critique of the author function, not necessarily as a dismissal of authorial criticism but as an anatomization of how an author’s name “points to the existence of certain groups of discourse and refers to the status of this discourse within a society and culture” (“What is an Author?” 123). A founding premise of the “discursive narratology” that I am attempting to elaborate is that a work of fiction is a public statement circulating in the same discursive formation as its author’s nonfictional statements. Furthermore, narratives are not static, for they are read differently each time according to their context of reception. Narrative authority, then, operates via a continuum between narrative voice, extrafictional voice, and authorial voice. [End Page 108] These voices have different textual forms and diegetic levels, but they coexist as public statements in the same discursive field and operate as interrelated rhetorical strategies for asserting the cultural significance of the novel to public life which establish a dialogue with the public response.

We can derive a theoretical framework for studying this continuum of voices from Genette’s concept of the paratext. For Genette, the “verbal or other productions” which frame and present a literary work to its readership comprise the paratext: “what enables a text to become a book and to be offered as such to its readers and more generally, to the public” (Paratexts 1).⁹ This paratext is a threshold between the text and its frame that “offers the world at large the possibility of either stepping inside or turning back” (2). Genette describes a spatial relationship between text and paratext, so that the location of a paratextual element “within the same volume” can be defined as a peritext, while the elements “located outside the book, generally with the help of the media,” can be defined as the epitext (4, 5). “In other words,” Genette writes, “for those who are keen on formulae, paratext = peritext + epitext” (5, emphasis original). Genette also posits a temporal relationship between text and paratext, with prior, original, and later or delayed paratexts defined in relation to the date of the text’s original publication.

The significance of the paratext to my approach lies in Genette’s emphasis on the pragmatic status of the text as a form of authorial communication in which the addressee is the public. “By definition,” Genette claims, “something is not a paratext unless the author or one of his associates accepts responsibility for it, although the degree of responsibility may vary” (9). Of most importance is Genette’s emphasis on the functionality of the paratext:

Indeed, this fringe, always the conveyor of a commentary that is authorial or more or less legitimated by the author, constitutes a zone between text and off-text, a zone not only of transition but also of transaction: a privileged place of a pragmatics and a strategy, of an influence on the public, an influence that—whether well or poorly understood and achieved—is at the service of a better reception for the text and a more pertinent reading of it (more pertinent, of course, in the eyes of the author and his allies.

(2) [End Page 109]

My approach has two crucial points of departure from Genette’s model. First is my claim that this paratext is the author’s attempt not only to frame a positive interpretation of the fiction, but to establish the fictional text as the basis for cultural authority as a public figure. For my purposes, Lanser’s extrafictional voice can be located within the peritextual elements of the book, linked via a discursive continuum to the authorial voice manifested in what Genette calls “the public authorial epitext,” composed of interviews, essays, etc. Secondly, I define this paratext more broadly than Genette in the sense that if it constitutes a “zone of transaction,” an attempt to influence the public, this zone also must include the textual phenomena produced by the reading public as the other party in the transaction. The interview, for instance, a key feature of the “public authorial epitext,” necessarily includes readerly responses in the form of the interviewer’s questions and itself exemplifies a transaction between author and reader over the significance of the text. The paratext, I am arguing, is a type of discursive formation, a set of textual statements whose interrelations construct the text as its object, which leads to a discursive reformulation of the diagram of narrative communication:

Click for larger view
View full resolution

A discursive approach to the narrative communication model situates the narrative text in a broader discursive formation to investigate how narrative authority emerges out of the relations between subject positions within this formation. So the epitext (author and reader), the peritext (extrafictional voice), and the text (narrative voice) contain the discursive sites at which these subject positions are articulated, and together the three sites constitute the paratextual zone of transaction, the discursive formation, in which what is being “transacted” is not so much textual meaning but the significance of the text to public discourse. I have excluded the implied author and implied reader from this model because, while they may be legitimate critical/cognitive constructs to facilitate reading, they are not concrete subject positions within or without the text so much as anthropomorphic postulations of the act of reading. I have retained the narratee as a [End Page 110] fictional subject position because, especially in omniscient narration, a specific narratorial address gives it textual form. The two-way arrows indicate that each discursive site facilitates a dialogue between the subject positions, that communication is always ongoing, drawing into play the temporal relations of a zone of transaction, and that the text itself gestures outwards or beyond to public dialogue on the paratextual level.

A key reason for postulating an implied author has been the need to retain the valuable theoretical distinction between author and narrator. Lanser points out, though, that if “an author-narrator separation is true in the abstract, it is nonetheless not abstractions that determine the reading of literature, but the conventions governing linguistic and literary use” (149). She goes on to argue that “in the absence of direct markings which separate the public narrator from the extrafictional voice, so long as it is possible to give meaning to the text within the equation author = narrator, readers will conventionally make this equation” (151). Omniscient narration is one fictional form for which such an equation is traditionally made, the recognition of which is crucial for understanding omniscient narrative authority as more than a function of literary convention; hence the need to frame the narrator-author relationship in pragmatic, flexible terms beyond formalist binaries if we are to understand the historical contingency of omniscient narration. The narrator and the author may be separate entities, but the act of narration, while fictional, is nonetheless a form of public discourse attributed to an author. The intrusive commentary of omniscient narration draws attention to this relationship, so that the “fictionality” of its discourse can be seen as a rhetorical device for asserting the importance of the novelist in public intellectual life, particularly when this narrative voice resonates textually with the extrafictional and extratextual voices of the author. As Lanser points out, “the equivalence of author and narrator implies an authorial responsibility that is similar to an author’s responsibility for his or her nonfictional work” (153).

So if omniscient authority is not so much a textual phenomenon, the narrator’s complete knowledge of the fictional world, as a type of narrative performance articulated through commentary, it gestures outwards, extratextually, to a particular figure of authorship. And since the narrative authority of contemporary omniscience no longer relies, as it did in previous centuries, on the consonance of its formal conventions with the cultural [End Page 111] authority of the novel itself, this performance must necessarily operate with a tension between its form and its status. The narrators of contemporary omniscience, I am suggesting, must gesture outwards to the broader realm of public discourse, in which less “universal” modes of public address circulate, in order to gain traction for their commentary. Knowledge of authorial voice consequently becomes important not for anchoring a biographical reading of a book, but for understanding how contemporary omniscient narration engages with the very question of novelistic authority.