I etcetera: on the poetics and ideology of multipersoned narratives - Second-Person Narrative

Brian Richardson

you have to say, It's I who am doing this to me, I who am talking about me to me. Then the breath fails, the end begins, you go silent, it's the end, short-lived, you begin again, you had forgotten, there's someone there, someone talking to you, about you, about him ... all I have to do is listen, then they depart, one by one, and the voice goes on, it's not theirs, they were never there, there was never anyone but you, talking to you about you, the breath fails (Beckett, The Unnamable 394)

One of the most significant omissions in contemporary narrative theory is the absence of sustained accounts of multiple narration. For the most part, narrative theory generally proceeds as if all novels were written entirely in the first person or third person, heterodiegetic or homodiegetic. One understands the desire for unambiguous foundations and paradigmatic examples, and naturally all theorists go on to identify more complex and ambiguous strategies of narration, but this situation leads to a conceptual framework in which the univocal practice is set forth as a (perhaps unintentional) norm while more heterodox experiments are consequently treated as secondary, peripheral, or even perverse literary gamesmanship. In this way, models of narrative ultimately derived from linguistics take precedence over Bakhtinian claims of the fundamentally polymorphous nature of the novel. Thus, while many typologies contain a space for both Bloom's subvocal speech and Molly's internal monologue, there is usually no place in such schemas for Ulysses as a whole as if the conjunction of different narrators and modes of narration were not itself of primary theoretical importance.(1) This gap is all the more unfortunate when one considers a work like The Sound and the Fury, in which the first-person "memory monologues," as Dorrit Cohn calls them (247-55), are starkly juxtaposed to the resolutely third-person segment that concludes the novel.

In what follows I will examine a number of different texts that employ multipersoned narration. Four major kinds of multipersoned texts may be identified at the outset: works that systematically oscillate between different narrative positions, those that collapse apparently different types of narration into a single voice, works whose narration remains fundamentally ambiguous, and texts that employ narrational stances that would be impossible in nonfictional discourse. Since many of the authors and critics of the works I will be discussing frequently invoke ideological reasons to explain their chosen narrative practice, a short excursus on the politics of narrative person will also be set forth. Throughout, I will pay particular attention to texts which employ second-person narration. It is no exaggeration to state that the emergence of second-person narration is precisely what makes speculation on multipersoned fiction inevitable. In a frequently quoted aside, Wayne Booth remarked in 1961 that perhaps the most overworked distinction in the theory of fiction is that of person (150). I believe we will find instead that person remains one of the most undertheorized distinctions in the field.

Alternating narration between different grammatical persons is not an exclusively modern phenomenon. As Stanzel has pointed out, Thackeray's narrator employs both "I" and "he" to describe his life in Henry Esmond. In Bleak House, an omniscient third-person narrative is juxtaposed to the first-person account of one of the characters. Stanzel explains that the "two narrative situations represent two different perspectives, namely, the panoramic one of the authorial narrator who is critical of the times and the naive but sympathetic viewpoint of the first-person narrator, Esther Summerson, circumscribed by her domestic horizons" (71). Even in this description, it might be noted, we get a sense of the ideological valences present in such a gendered division of knowledge and narration. As Susan Sniader Lanser points out, by "replicating the ideology of separate spheres," Bleak House sets "the omniscient and implicitly male voice of the authorial narrator next to the personal voice of the female character Esther Summerson without acknowledging this duality" (239-40). I will return to the subject of the poetics of gender and the politics of pronouns later in this essay; for now we can conclude this historical preface by mentioning other titles Stanzel cites: Conrad's Under Western Eyes, Robert Penn Warren's All the King's Men, Bellow's Herzog, and several novels by Max Frisch, though even here it is necessary to point out an additional type of narration that Stanzel neglects to mention: the second-person narrative that constitutes the opening section of All the King's Men.

Contemporary fiction is replete with a polyphony of competing narrative voices; even where the narrator's speaking situation seems fixed, the proliferation of alternative voices threatens to destabilize that situation. A representative example of this kind of play with person can be found in Nathalie Sarraute's Tu ne t'aimes pas (1989). The novel begins as a dialogue between the different figures bound together within a single self, as the following passage discloses:

Vous ne vous aimez pas? Qui n'aime pas qui?

- Toi, bien sur . . . c'etait un vous de politesse, un vous qui ne s'adressait qu'a toi.

- A moi? Moi seul? Pas a vous tous qui etes moi . . . et nous sommes un si grand nombre . . . "une personnalite complexe" . . . comme toutes les autres. . . . (9)

You don't love yourself? Who doesn't love whom? You, of course . . . you, the only one they were talking to. Me? Only me? Not all the rest of you who are me? . . . and there are so many of us . . . "a complex personality" . . . like every other. . . . (1)

This unusual colloquy continues, "tu t'es separe de nous, tu t'es mis en avant comme notre unique representant . . . tu a dit <<je>> . . ." (9) ("you broke away from us, you put yourself forward as our sole representative . . . you said 'I' . . .") (2); two lines later, an additional twist is suggested by the claim "[c]haque fois que l'un de nous se montre au-dehors, il se designe par <<je>>, par <<moi>> . . . comme s'il etait seul, comme si vous n'existiez pas . . ." (9-10) ("Every time one of us shows himself to the outside world he designates himself as 'I,' as 'me' . . . as if he were the only one, as if you didn't exist . . .") (2).

Passages like these effectively dramatize the intersubjective constitution of "the" self, the instability of the classical ego, and pose interesting questions about gender, speech, and subjectivity (who is the "il" that suddenly evolves?). Both "vous" and "tu" are evoked here although just what these terms refer to is by no means obvious. One can also find a more realistic though equally ambiguous separation and conflation of person and voice in "Unguided Tour," the last story in Susan Sontag's collection I, etcetera (1978), in which an apparent dialogue between an "I" and a "you," transcribed without quotation marks, is collapsed into other voices and texts.

Similarly, Clarice Lispector's Agua Viva (1978) seems to be a missive from a woman to her lover with "I" and "you" firmly locked in their conventional places. As the narrative continues, the "I" disintegrates: "I divide myself thousands of times, into as many times as the seconds that pass, fragmentary as I am and precarious the moments" (4). While the speaking self is continuously dispersed and multiplied, the addressee expands to include the various readers of the text: "I write you completely whole and I feel a pleasure in being and my pleasure in you is abstract, like the instant" (4). These two pronouns and their elusive referents are soon further reconstructed, as the narrator announces that "if I say 'I,' it's because I don't dare say 'you,' or 'we,' or 'a person.' I'm limited to the humble act of self-personalmzation through reducing myself, but I am the 'you-are"' (6). Here, an act of apparent humility - not presuming to speak for someone else - is simultaneously disclosed to be a bold claim to do just that. And this occurs through the curious act of "self-personalization through reducing myself," a practice that superficially can seem both tautological and self-contradictory unless we recognize it as a persuasive statement of the intersubjective constitution of a fluid and multiform subjectivity.

Lispector's drama of person, self, and other continues as the narrator explains that she is "still not ready to speak of 'him' or 'her"' (28); the introduction of the third-person pronoun proper in the form of the story of Joao will not be presented until much later in the text (48). In the meantime, the narrator is transfixed by the pursuit of the "it," as in "I need to feel the it of the animals again" (38), that is, a kind of primordial sensibility beneath or beyond mere individuality: "it seems I'm achieving a higher plane of humanity. Or of inhumanity - the it" (43). At this point, other pronouns fuse: "You have become an I" (43). The work ends with a pseudoclassical resolution. The narrator, who began by asking "Who am I?" claims to have found herself, and in the process realizes the disjunctive corollary that, after all, "you are you" (79). But it is no longer clear exactly what these terms refer to after their conventional meanings have been so thoroughly effaced. The narrative practice of Sarraute, Sontag, and Lispector outlined above tends to elude conventional theoretical models because it implodes or transcends the stable, determinate identities presupposed by those theories.

The importance of investigating the pronouns that narrators employ to designate themselves is evident from the number of narratives whose plot hinges upon the determination of the narrator's identity. In "The Shape of the Sword" (1944), Borges's protagonist recounts, in the third person, the story of a cowardly traitor; only by the end of the tale do we realize that the narrator has been describing himself all along.(2) Elsewhere, I have termed this practice a "pseudo third-person" narrative and identified a dramatic analogue in Beckett's short play Not I ("Drama" 202-04). Similarly, Didier Husson has recently drawn attention to Camus's unusual fusion of heterodiegetic and homodiegetic narration in La Peste, where the seemingly heterodiegetic narrator is revealed in the end to be the principal character, Dr. Rieux. The same strategy also appears in Robbe-Grillet's Dans le labyrinthe (1959) though here its abrupt and seemingly gratuitous application may very well be a parody of Camus's text. Other interesting variants of this play with narration include Borges's "The Immortal" (1949), where the narrator of the unlikely first person inner story (who claims to have spoken to Homer and drunk from a fountain of youth) discovers that he is not merely unreliable, but impossible: "The story I have narrated seems unreal because in it are mixed the events of two different men" (117): that is, the narrator, in recounting his life story, has conflated it with another's. The framing material at the end of the text resolves this anomaly: the tale does not, in fact, entwine the lives of disparate individuals within the spurious identity of a single "I." Instead it is revealed that the narrator is simply a deluded fabricator, writing a fiction in the first person that he originally believes to be an autobiography and then equally erroneously supposes is the story of an "I" and a "he."

Two other texts that contain abrupt revelations or complications of the true identity of the narrator might also be mentioned. Calvino's Il cavaliere inesistente (1959) is, at the beginning, a third-person narrative of the postmodern adventures of several of Charlemagne's paladins, including Raimbaut's pursuit of the amazonian warrior Bradamante. Some thirty pages into the novella, however, the author of the narrative makes her presence known and reveals herself to be a nun in a convent who is trying to imagine the foreign scenes she recounts. At the end of the text she again employs the first person and discloses that she is in fact the very Bradamante she had been describing in the third person and as the object of a male quest. The strategy of narration thus tends to negate the ideological valence implicit in the story's teleology, that of the (con)quest of the female.

In Joyce Carol Oates's story "You" (1970), we are presented with what I would call a pseudo second-person narrative. The first third of the story seems to be a straightforward example of what I have elsewhere termed standard second-person narration ("Poetics"), in which the protagonist, focalizer, and governing consciousness of the text is a single figure designated by the second-person pronoun. This assumption is strengthened by narration that seems to depict the stream of thoughts of the titular character: "It strikes you that this is an important scene, an emotional scene. People are watching you anxiously. You might be in a play. Not one of those crappy television plays . . ." (365). A little more than a third of the way into the story, however, a first-person narrator emerges; it turns out that the tale's narrator is the neglected teenage daughter, and the "you" refers entirely to the daughter's imaginative construction of the probable events, physical and mental, of her mother's life.(3) Once again, the plot of the narrative and the poignancy of the events turn on the revelation of the narrator's actual identity, as the text's play with the conventions of narration and its reception is disclosed.

Recently, a number of novels have appeared that alternately use the first-and the third-person pronouns to designate the same character. A survey of some of these works may help establish precisely what may be at stake in the deployment of multiple personal reference, both aesthetically and ideologically. The first chapter of Fay Weldon's The Cloning of Joanna May (1989) is narrated in the first person and is followed by three third-person chapters; then the work returns to the first person for a chapter that is followed again by three more third-person chapters. This pattern, with a few basic variations, continues to the end of the novel. Since the first-person account is that of a woman struggling to achieve a sense of self and escape the depredations of the brutal patriarchal world, which threatens first to marginalize and then to kill her, we may be forgiven if we first perceive this text as another reinscription, pace Dickens, of the public (male)-private (female) hierarchy. In fact, however, a dual epistemology is presented in the text: one deterministic, materialistic, patriarchal, and murderous and the other more spiritual, communal, female, and alert to the play of chance. The ultimate triumph of Joanna May and her achievement of self, community, and survival - the victory, that is, of the "I" over the "he" - constitutes a rewriting and reversal of the very hierarchy inscribed by Dickens. Intriguingly, many of the novel's thematic oppositions are first expressed in terms of the second person:

"Love" I could understand, but what did he mean by this "you"? Small children (so I'm told) start out by confusing "me" with "you." Addressed so frequently as "you," their clever little minds work out that this must be their name. "You cold," they say, shivering, as the wind blows through the window. "Not you," comes the response, "me." "Me cold," says the child, obligingly. Presently the little thing progresses to the gracious "I am cold." But is the "me," the "I," really the same as that initial "you" with which we all begin; the sudden bright consciousness of the self as something defined by others? Perhaps we did better in our initial belief, that the shivering cold is jointly experienced, something shared. I wonder. (6)

Other contemporary works articulate comparable notions. In Michel Tournier's Friday (1967), as Arlette Bouloumie has pointed out, a past-tense, third-person account alternates with a first-person logbook, reflecting the thematic conflict between the man Crusoe was and the person he is becoming, as one form contaminates the other and results in "a triumph of the logbook and a victory for interiority" (Bouloumie 454). Marguerite Duras's switching between the first- and third-person pronouns in L'amant (1984), on the other hand, suggests a more profound, reflexive, and unsettling narrative topography. As Sharon Willis explains, "given the text's strategy of veiling and unveiling, where 'I' veils herself as 'she,' but where 'she' just as frequently masquerades as 'I,' we cannot maintain a rigid and secure separation of self and other, interior and exterior. Nor can we as readers determine a fixed vantage point, and the reassuring distance that would entail" (6).

Christa Wolf, whose Nachdenken uber Christa T. (1970) deftly mingles first- and third-person accounts, alternates between second- and third-person passages in Kindheitsmuster (1976) as the "I" is avoided and displaced by a "you" ("du"). As the narrator explains: "Allmahlich, uber Monate hin, stellte sich das Dilemma heraus: sprachlos bleiben oder in der dritten Person leben, das scheint zur Wahl zu stehen. Das eine unmoglich, unheimlich das andere" (9) ("Gradually, as months went by, the dilemma crystallized: to remain speechless, or else to live in the third person. The first is impossible, the second strange") (3). At this point, the "unheimlich" second-person narration commences. Throughout the novel, the second person is presented as more authentic, more responsible, and perhaps less subjective than the first person, though less arduous than the third.

In Roland Barthes's book Roland Barthes (1975), we find a different oscillation, one between "I" and "you" segments, which also includes leading comments concerning the use of pronouns and the nature of writing one's life. ("All this must be considered as if spoken by a character in a novel - or rather by several characters" [119]; "The intrusion, into the discourse of the essay, of a third person who nonetheless refers to no fictive creature, marks the necessity of remodeling the genres: let the essay avow itself almost a novel: a novel without proper names" [120]). By writing autobiographical fragments with techniques borrowed from contemporary fiction, Barthes not only problematizes the notion of genre, but also produces a Borgesian frame that keeps turning on itself. Andrew Brown observes of this passage that "[s]omebody in Barthes par luimeme is saying il, framing Barthes: that somebody is a fictional character . . . but the il is not a creature fictive: Barthes has thus ensured that in such passages he has written a fiction whose (unnamed and unreal) soliloquist does nothing but discuss the (real) il" (124).

Additional examples of this and analogous practices are plentiful; one might point to Nadine Gordimer's The Conservationist (1974), Sherley Anne Williams's Dessa Rose (1986), J. M. Coetzee's Foe (1986), Marge Piercy's He, She and It (1991), and Carlos Fuentes's Cambio de piel (1967), the last named work using second- and third-person narration. Together, these texts reveal the following distinctive characteristics of multipersoned narrative: fresh possibilities for formal literary innovation; new methods to reinscribe thematic material at the level of narration, as the text's central concerns are embodied in a correlative formal technique; alternative tools to define more sharply or collapse more effectively conventional distinctions between different characters, competing narrative worlds, tale and frame, or text and reader; and cunning embodiments of contemporary issues in philosophy, cultural studies, and gender theory.

All of these features figure even more prominently in the small group of unusually resonant fictions that juxtapose first-, second-, and third-person narration. Fuentes uses all three persons alternately to reveal the sweeping transformation of the identity of the protagonist (and, by extension, of the Mexican revolution and its aftermath) in La muerta de Artemio Cruz (1962). More radically, Juan Goytisolo's Paisajes despues de la batalla (1982) ingeniously manipulates reader identification and response through a deviously compelling play with narrative person. The shifting use of the narrative "you" is used to establish an identity between the reader and the dispossessed peoples living in and around Paris as well as with a xenophobic racist and child molester. In Calvino's Se una notte d'inverno un vaggiatore (1979), a multifaceted second-person text (that generates a number of distinct second-person figures) frames a series of ostensibly independent first-person narratives along with occasional third-person segments.(4) In this book, an expanding number of ever more artificial characters and plots serve as a foil for both the several "you's" and a chameleonic "I."

A more systematic interrogation of the three major narrational pronouns appears in Nuruddin Farah's Maps (1986). As Rhonda Cobham points out, the shifting uncertainties of nation, sexuality, and history are mirrored in the text's "use of three personal pronouns to narrate Askar's story. Askar identifies these at the end of the novel as the voices of judge, witness, and audience (246), although, typically, the lines between these three perspectives are not always reliably indicated by the pronoun used" (49). Each narrative voice both obscures and reveals different information and does so in an idiosyncratic tone. The first-person narrative is rather diffident and at times almost wilfully ignorant of the surrounding circumstances; the second-person narration is on the other hand vague, oneiric, vatic, and suspiciously profound. The third-person narration, though not without its own archetypal ambiance, is more objective, limiting, and distant. It is precisely these differences that foreground the paramount significance of the category of person in the narrative transaction and suggest the conclusion that any narrative theory ignoring such distinctions - so crucial to the practice of the writers just discussed - must remain seriously impoverished.

Genette states: "The novelist's choice, unlike the narrator's, is not between two grammatical forms, but between two narrative postures (whose grammatical forms are simply an automatic consequence): to have the story told by one of its 'characters,' or to have it told by a narrator outside of the story" (244). But this is precisely what Farah's Maps refuses to do. The second-person narration occupies a middle position, oscillating between but never reducible to either the homodiegetic or the heterodiegetic. Many passages suggest that a narrator-protagonist is referring to himself in the second person: "Alone, again, once you knew how to write your name, you would secretly graft your name, born the same day as the tree, on its bark" (63-64). At other points however the protagonist appears to be depicted externally: "You sit, in contemplative posture, your features agonized and your expression pained . . ." (1).(5) A similar ambiguity attends the novel's three-personed narrative form. Unlike Bleak House, in which Esther Summerson's first-person account can be neatly placed within a larger, extra- and heterodiegetic whole, the three narrations of Maps refuse to fall into any epistemological hierarchy. We cannot determine whether Askar is telling his story using three different pronouns or whether an extra- and heterodiegetic narrator is employing all three forms or whether two or three distinct narrators are at work. It is precisely this irreducible ambiguity that gives the novel its peculiar tension and urges theorists to reconceive and extend our analytical categories.

This conclusion is perhaps even more evident when one considers Maurice Roche's Compact (1966), probably the ne plus ultra of multipersoned narrative. Its multiple, symmetrical, and curiously parallel story lines are narrated in a nearly exhaustive range of possible pronouns that foreground the insistence and complexity of the category of narrative person. The reader is presented with you (tu), one (on), I, he, we, and even a passive voice narration devoid of pronominal reference.(6) The power of the alternating narrations is often most evident in transitions from one form to another:

J'avais beau me dire que claquer etait a la portee de tout le monde, que ce serait une bonne chose de faite, le coeur n'y etait pas.

"-- . . . humble medecin est un hypnotiseur parfait; au besoin, il utilise protoxide d'azote . . . operation benigne . . ." IL S'AGISSAIT DE LUI EPARGNER TOUTE SOUFFRANCE.

Tu ne sentiras rien. Tout se passera a ton insu; tu pourrais aussi bien ne pas etre la; mais loin, hors de tout. (70)

I could tell myself a million times that turning it in was within everyone's reach, that it would be a deed well done, my heart just wasn't in it.

"-- . . . humble doctor is perfect hypnotist; if necessary, he uses nitrous oxide . . . benign operation . . ." IT WAS A MATTER OF SPARING HER PAIN.

You'll feel nothing. Everything will happen without your knowing it. You could just as well not be there, but far away, outside of everything. (58)

The meticulous, contrapuntal braiding of distinct voices (and the independent stories they narrate) in this text clearly indicates person to be an indispensible category of narrative analysis. The fact that, as many critics have observed, some of the different narrative voices seem to merge together testifies to the need for particular suppleness within such an analytical perspective.(7)

In the history of narrative theory we can find an intriguing precedent for the position I am advocating. In 1927, E. M. Forster attacked the univocal categories propounded by Percy Lubbock and praised instead Bleak House and Gide's Les Faux monnayeurs, works in which omniscient narration, limited omniscience, and first-person forms alternate. Forster concluded:

A novelist can shift his view-point if it comes off, and it came off with Dickens and Tolstoy. Indeed this power to expand and contract perception (of which the shifting view-point is a symptom), this right to intermittent knowledge: - I find it one of the great advantages of the novel-form, and it has a parallel in our perception of life. (81)

It is unfortunate that this thesis of Forster's has been critically overlooked for so long by Lubbock's varied followers.

The second point Forster makes in this passage - the relation between the strategies of narration and lived social experience - has drawn a great deal more attention, particularly in the numerous quests not merely to identify general ideological tendencies but to equate one form of narration with a specific political agenda. For much of this century, theorists have debated the ideological valence of the choice of person in narration. Glancing back over some of these controversies, one is struck by the vehemence of the argumentation, the abrupt shifts in position over what is "revolutionary" and what "reactionary" from decade to decade, and the general ephemerality of moments of consensus: does anyone still take Lukacs's strictures against modernist subjectivism seriously? Does anyone still care about the once celebrated Sartre-Mauriac controversy?

Roland Barthes is perhaps the most persuasive of recent ideological apologists. Developing a position first set forth in Writing Degree Zero, Barthes states that "'he' is wicked: the nastiest word in the language: pronoun of the non-person, it annuls and mortifies its referent; . . . Saying 'he' about someone, I always envision a kind of murder by language . . ." (Roland Barthes 169). Though more sophisticated and paradoxical than his earlier simple denunciation of third-person narration, Barthes's suspicion of the third person remains clear.(8) While we may sympathize with the general position behind his remarks - that third-person omniscient narration tends to reify and "naturalize" existing social relations, the fact remains that, among writers who alternate from book to book between first- and third-person narratives (Dickens, Thackeray, Conrad, Calvino, Wittig), we do not usually find the first-person narratives more emancipatory than the others; sometimes, indeed, we do not immediately remember whether a given text is narrated in one person or the other. Furthermore, many works of certain avowed fascists and their sympathizers (Hamsun, Celine, Cela) employ a particularly intense kind of first-person narration. These examples would seem definitively to refute any claims for the liberatory nature of first-person narration. One suspects instead that no form has any inherent essence or tendencies - or at least none that a competent practitioner cannot circumvent. Ideological stances are frequently enmeshed with practices of narration, but never in a way that can be reduced to an easy equation.

The question of gender and person is more complex and raises somewhat different issues. Joanne S. Frye, in a chapter on the "subversive I," states that for female writers, to "speak directly in a personal voice is to deny the exclusive right of male author-ity implicit in a public voice and to escape the expression of dominant ideologies upon which an omniscient narrator depends" (51). She goes on to explain that:

If a female pronoun recurs throughout a text it repeatedly reminds us of cultural expectations for what it means to be female; it reminds us, inevitably, of the [patriarchy's] femininity text. The "I," by contrast, reminds us only of a subjective narrating presence, a nameless agent; it asks us to remember only its subjective agency. The "she" can easily lull us into conventional expectations; the "I" keeps us conscious of possibility and change. (65)

This value-laden binary opposition, at once suggestive and familiar, necessarily invites scrutiny of the terms it rests upon. First of all, it tends to replicate rather than contest what Lanser, in her discussion of Bleak House cited earlier, identified as "the ideology of separate spheres" (240), a division many feminists might well be reluctant to see codified so decisively. In addition, many might question the subversive or liberatory value posited in the first-person pronoun. Virginia Woolf, it will be remembered, quickly became bored by a representative male writer's insistent use of the first person: "One began dodging this way and that to catch a glimpse of the landscape behind it. Whether that was indeed a tree or a woman walking I was not quite sure. Back one was always hailed to the letter 'I'. One began to be tired of 'I'" (103). Far from erasing or eluding gender markers, the "I" of the male writer is utterly self-absorbed and altogether too self-sufficient.(9)

The feminist quest for a discursive site beyond the somewhat predictable first person is most emphatically articulated by Monique Wittig in the author's note that prefaces Le corps lesbien (1975). She states that the "je," when written by a woman, is always alienating since that "I" must write in a language that denies and negates female experience. Every such usage is always already reinscribed within a larger masculine matrix (10). All of the limitations Frye attributes to the use of "she" are present for Wittig in the writing of "je." Consequently, Witrig says that she is physically incapable of writing je; instead, her novel is narrated by a j/e, an ideologically marked pronoun that desires "to do violence by writing to the language which I [j/e] can enter only by force" (10).(10)

In her essay "The Three Genres," Luce Irigaray makes an important observation in her analysis of male and female pronominal usage, the first part of which might serve as a gloss on Woolf's remarks: "With men, the I is asserted in different ways; it is significantly more important than the you and the world. With women, the I often makes way for the you, the world, for the objectivity of words and things" (146). Women's discourse, even in the first person, often remains other-directed, seeking interaction and validation. Men, on the other hand, "live within the closed universe of the first-person pronoun; their messages are often self-affirmations which leave little for co-creation with an other sex," writes Margaret Whitford, summarizing Irigaray's position (78). This postulate can help explain why a number of feminist writers have moved beyond the I to the more fluid, interactive, and destabilizing technique of second-person narration found in works like Mary McCarthy's "The Genial Host," Edna O'Brien's A Pagan Place, Lorrie Moore's Self Help, Jamaica Kincaid's A Small Place, Pam Houston's "How to Talk to a Hunter," and Sunetra Gupta's The Glass-blower's Breath - in fact, feminist second-person narration can be traced back to the opening sections of May Sinclair's Mary Olivier (1919). This postulate also suggests specific ideological reasons for other feminists' deployments of still other pronouns (most strikingly, on and elles in Wittig's L 'Opoponax [1964] and Les Guerilleres [1969]) and for the alternation and/or conflation of the first-, second-, and third-person pronouns in previously mentioned works by Lispector, Sarraute, Sontag, Wolf, Weldon, and Duras. One may, in conclusion, applaud the efforts of feminist writers (and, by extension, those of sexual and ethnic minorities) to overcome by means of innovative strategies of narration undeniable sociological prejudices inherent in existing linguistic practice, even as one acknowledges certain difficulties in establishing the exact parameters of such prejudices or the precise efficacy of specific interventionary strategies.(11)

There is one final category of multiple narration to identity, and that is the category of "impossible narration": texts containing discourse that cannot logically be spoken or written by their purported narrators and usually involving the kind of ontological frame breaking typical of postmodern works. A paradigmatic example of this kind of text is Christine Brooke-Rose's Thru (1975), which is succinctly described by Shlomith Rimmon-Kenan in the following terms: "The novel repeatedly reverses the hierarchy [of narrative levels], transforming a narrated object into a narrating agent and vice versa. The very distinction between outside and inside, container and contained, narrating subject and narrated object, higher and lower level collapses, resulting in a paradox which the text itself puts in a nutshell: 'Whoever you invented invented you too'"(94).

This practice, which can only occur in works of fiction, extends back at least as far as Diderot's Jacques le fataliste and is not uncommon in recent French fiction, notably Beckett's The Unnamable and the later novels of Robbe-Grillet. The Unnamable, it might be observed, simultaneously enacts several of the different possibilities of multipersoned narrative outlined above. The narrative voice often refers to other characters and voices, which it then discloses to be fictions invented by itself, as apparently independent persons are collapsed into a single narrative voice. On the other hand, the same narrator goes on to claim to have invented the frustrated narrators of other novels written by Beckett; the narrator of The Unnamable is here impersonating its author. Throughout the text, the frequently used pronoun you designates varied and ambiguous figures. This narrative strategy does not merely problematize conventional theoretical distinctions, but undermines the very terms upon which such distinctions rest.

Traditional narrative theory, perhaps implicitly based on the more stable nonfictional types of biography (third person) and autobiography (first person), has a difficult time comprehending forms that, like second person and impossible narration, do not or cannot occur in nonfictional narrative. Theory routinely tends to deny, ignore, or dismiss as inconsequential curiosities the very kinds of narration that are distinctively fictional and that as such would seem to demand inclusion in a comprehensive poetics of fiction, all the more so since the history of fiction is rife with similar if brief examples even in ostensibly stable, familiar, and homogeneous texts: consider the "found manuscript" episode of Don Quixote or Proust's few though notorious acts of frame breaking in the Recherche.(12) The fundamental question is, once again, which model of narrative is more effective for theorizing the practice of fiction: one grounded in linguistics and imitations of nonfictional narratives or one that begins with the heterogeneity, polymorphism, and flagrant fictionality typical of the novel from Petronius and Lucian to Beckett and Wittig?

If, as I have argued, the latter is the case, it is necessary to remap strategies of narration in a more expansive and dialectical manner. Remembering Stanzel's "narrative circle," which uses a subtly gradated dyadic illustration of narrative person (first and third person, naturally), we can now envisage an alternative model that is more comprehensive and flexible. I would like to argue for a four-part division of such a circle. [ILLUSTRATION FOR FIGURE 1 OMITTED] The quadrant on the right would cover first-person narration; its opposite, on the left, would include third-person forms. At the bottom, connecting the two, free indirect speech can be situated; at the top, the long neglected category of second-person narration can take its rightful place. Second-person novels that incline toward other forms of presentation may be situated accordingly: Jay McInerney's Bright Lights, Big City would be near the first-person boundary, while Edna O'Brien's A Pagan Place should reside near the space of the third person.

We may further subdivide this quadrant to include a number of recent innovative deployments of person.(13) Texts using a narrative "we," such as Pierre Silvain's Les Eoliennes (nous), belong for obvious reasons at the boundary between "I" and "you," just inside the first-person division. "One" narration (on, man), which nearly approximates much of the same conceptual space as "you" forms, will be the third-person form closest to the second-person boundary. On the other side of this intersection between second- and third-person narration we may situate the "subjunctive" second-person or "recipe" mode, in which a narratee-protagonist is given instructions that form a narrative (John Updike's "How to Love America and Leave It at the Same Time" in Problems [1978]).(14) In addition, the category of the third person can be reconceptualized to include works that refer largely to a "they" (Perec's Les Choses [1965]), an "it" (briefly in Isherwood's A Single Man [1964] and Fowles's Mantissa [1982]) and "na," the gender-neutral pronoun invented by June Arnold for her novel The cook and the carpenter (1973). (It is interesting to note that, in outlining a thorough delineation of formally innovative strategies of narration, one continues to encounter highly charged ideological positions.)

To conclude our mapping of narrational possibilities, it will be helpful to draw another circle around the first circle to represent the flow of multipersoned narratives that travel across or beyond the ordinary divisions and juxtapose divergent points of view. Lastly, I would like to propose that a large black dot occupy the center of the figure to stand for impossible and impossibly conflated acts of narration. Only with such a Borgesian schema can one embrace the kinds of fictions produced by authors like Borges.

In the preceding pages I have attempted to indicate some of the range, variety, and significance of multipersoned narrative. This long-established and increasingly popular form has been a vehicle for cunning aesthetic innovations, substantial philosophical investigations, and vigorous ideological affirmations. The collective richness of these works and their challenging effects on readers demonstrate convincingly the importance of person as a category of narrative theory and analysis even as they show that the concept of person must be substantially expanded to include second person, multipersoned, and "impossible" narration. This conclusion in turn suggests that the poetics of fiction is incommensurable with theories of nonfictional narrative: any thorough, systematic, universal narratology must do justice to the radical heterogeneity, ontological conflations, and logical impossibilities that proliferate in, and only in, fictional narratives. In other words, a genuinely comprehensive narrative poetics must include, as it were, its own negation and embrace works that fit smoothly within standard typologies as well as those that defy and transgress the typological imagination.

Notes

1 Two important partial exceptions to this practice are Franz K. Stanzel's A Theory of Narrative, which includes a fine discussion of alternating first- and third-person pronominal reference in, for example, novels written by a protagonist that at times refers to himself in the third person, as in Henry Esmond (99-110); and Hazard Adams points toward a more fluid model of the narrative transaction in "Critical Constitution of the Literary Text: The Example of Ulysses." Gerard Genette also describes several interesting examples of alternating persons in Narrative Discourse (243-47), only to deny the importance of person as a category of narrative analysis.

2 Genette discusses this example in Narrative Discourse (246-47) only to conclude, somewhat perversely, that "[t]he Borgesian fantastic, in this respect emblematic of a whole modern literature, does not accept person" (247); the entire plot of the story, which is the plot of its narration, affirms instead the paramount importance of person.

3 For an excellent narratological analysis of this story, see Monika Fludernik, "Second."

4 A number of these second-person identities are established by James Phelan (133-62).

5 For an extended discussion of this oscillation, typical of second-person narrative, see my essay on the subject (Richardson, "Poetics" esp. 311-19).

6 It is of course important not to confuse pronominal usage and narrative person. Thus an autobiographical account like Julius Caesar's Gallic War remains a first-person text even though the author describes himself in the third person. Similarly, as Fludernik has pointed out, a second-person narrative can employ a third-person pronoun such as the German "Sie" ("Second" 219-20).

7 For a compelling overview of the book's narrative stances, see Susan Rubin Suleiman (44-49).

8 For a perspicacious account of the complexities of Barthes's position, some of which are necessarily slighted in my summary remarks, see Brown (123-25). In Writing Degree Zero, Barthes had affirmed that "'he' is a typical novelistic convention," while the "I" can take "its place beyond convention" (35).

9 On the other hand, Frye's position seems to be partially corroborated by Maxine Hong Kingston's observation: "There is a Chinese word for the female I - which is 'slave.'" Kingston affirms that such a usage attempts to break "the women with their own tongues" (56).

10 For a more extended discussion of this pronominal strategy, see Erika Ostrovsky (76-78).

11 For additional discussion of how women and gays have used the second person and other uncommon pronominal forms to combat stereotyping and enhance potential reader identification, see Fludernik, "Pronouns."

12 The most abrupt instance is the authorial declaration near the end of the book that all of the characters are inventions except for the millionaire cousins of Francoise.

13 I am greatly indebted to Monika Fludernik and her article on "odd third person forms" for many of the examples that follow.

14 Much subjunctive second-person narration could be rewritten using one instead of you with little change in meaning, as the following sentence suggests: "To get there you follow [one follows] Highway 58. . . ." This similarity may have led Wittig's translator to use the English word you to render the ubiquitous on of L'Opoponax. Wittig's subsequent annoyance over this choice suggests an important difference between the two, a difference I suspect is rooted in divergent narrative persons.

Works Cited

Adams, Hazard. "Critical Constitution of the Literary Text: The Example of Ulysses." Antithetical Essays in Literary Criticism and Liberal Education. Tallahassee: Florida State UP, 1990. 90-110.

Barthes, Roland. Roland Barthes. Trans. Richard Howard. 1975. New York: Hill, 1977.

-----. Writing Degree Zero and Elements of Semiology. Trans. Annette Layers and Colin Smith. 1953, 1964. Boston: Beacon, 1968.

Beckett, Samuel. "Not I." Collected Shorter Plays. New York: Grove, 1984.

-----. Three Novels: Molloy, Malone Dies, The Unnamable. New York: Grove, 1965.

Booth, Wayne C. The Rhetoric of Fiction. 2nd ed. 1961. Chicago: U of Chicago P, 1983.

Borges, Jorge Luis. Labyrinths. New York: New Directions, 1964.

Bouloumie, Arlette. "Writing and Modernism: Michel Tournier's Friday." Style 26 (1992): 447-56.

Brown, Andrew. Roland Barthes: The Figures of Writing. Oxford: Oxford UP, 1992.

Cobham, Rhonda. "Misgendering the Nation: African Nationalist Fictions and Nuruddin Farah's Maps." Nationalisms and Sexualities. Ed. Andrew Parker, et al. New York: Routledge, 1992. 42-59.

Cohn, Dorrit. Transparent Minds: Narrative Modes for Presenting Consciousness in Fiction. Princeton: Princeton UP, 1978.

Farah, Nuruddin. Maps. New York: Pantheon, 1986.

Fludernik, Monika. "Pronouns of Address and 'Odd' Third-Person Forms: The Mechanics of Involvement in Fiction." New Essays on Deixis. Ed. Keith Green. Forthcoming.

-----. "Second-Person Fiction: Narrative You as Addressee and/or Protagonist." AAA - Arbeiten aus Anglistik und Amerikanistik 18 (1993): 217-47.

Forster, E. M. Aspects of The Novel. New York: Harcourt, [1927].

Frye, Joanne S. Living Stories, Telling Lies: Women and the Novel in Contemporary Experience. Ann Arbor: U of Michigan P, 1986.

Genette, Gerard. Narrative Discourse: An Essay in Method. Trans. Jane E. Lewin. 1972. Ithaca: Cornell UP, 1980.

Husson, Didier. "Logique des possibles narratifs." Poetique 87 (1991): 289-313.

Irigaray, Luce. The Irigaray Reader. Ed. Margaret Whitford. Oxford: Blackwell, 1991.

Kingston, Maxine Hong. The Woman Warrior. 1976. New York: Random, 1977.

Lanser, Susan Sniader. Fictions of Authority: Women Writers and Narrative Voice. Ithaca: Cornell UP, 1992.

Lispector, Clarice. The Stream of Life. Trans. Elizabeth Lowe and Earl Fritz. Minneapolis: U of Minnesota P, 1989.

Oates, Joyce Carol. The Wheel of Love and Other Stories. New York: Vanguard, 1970.

Ostrovsky, Erika. A Constant Journey: The Fiction of Monique Wittig. Carbondale: Southern Illinois UP, 1991.

Phelan, James. Reading People, Reading Plots: Character, Progression and the Interpretation of Narrative. Chicago: U of Chicago P, 1989.

Richardson, Brian. "The Poetics and Politics of Second-Person Narrative." Genre 24 (1991): 309-30.

-----. "Point of View in Drama: Diegetic Monologue, Unreliable Narrators, and the Author's Voice on Stage." Comparative Drama 22 (1988): 193-214.

Rimmon-Kenan, Shlomith. Narrative Fiction: Contemporary Poetics. London: Methuen, 1983.

Roche, Maurice. Compact. Paris: Seuil, 1966.

-----. Compact. Trans. Mark Pollizotti. 1966. Elmwood Park: Dalkey Archive, 1988.

Sarraute, Nathalie. Tu ne t'aimes pas. Paris: Gallimard, 1989.

-----. You Don't Love Yourself. Trans. Barbara Wright. New York: Brazilier, 1990.

Sontag, Susan. I, etcetera. New York: Farrar, 1978.

Stanzel, Franz. A Theory of Narrative. Trans. Charlotte Goedsche. 1979. Cambridge: Cambridge UP, 1984.

Suleiman, Susan Rubin. Subversive Intent: Gender, Politics, and the Avant-Garde. Cambridge: Harvard UP, 1990.

Weldon, Fay. The Cloning of Joanna May. 1989. New York: Viking, 1990.

Whitford, Margaret, ed. Introduction. The Irigaray Reader. By Luce Irigaray. Oxford: Blackwell, 1991.

Willis, Sharon. Marguerite Duras: Writing on the Body. Urbana: U of Illinois P, 1987.

Wittig, Monique. The Lesbian Body. Trans. David Le Vey. 1973. Boston: Beacon, 1986.

Wolf, Christa. Kindheitsmuster. Berlin: Aufbau, 1978.

-----. A Model Childhood. Trans. Ursule Molinaro and Hedwig Rappoll. 1978. New York: Farrar, 1980.

Woolf, Virginia. A Room of One's Own. London: Harcourt, [1929].

Brian Richardson is assistant professor of English at the University of Maryland. His articles on narrative theory and criticism have appeared in Genre, Style, Essays in Literature, Comparative Drama, Poetics Today, and The Faulkner Journal.

COPYRIGHT 1994 Northern Illinois University
COPYRIGHT 2004 Gale Group